March 18, 2025 • 44 mins
The Bloomberg Intelligence small-cap BMVP multi-factor portfolio outperformed the Russell 2000 equal-weight index last year, with value, momentum, low volatility and profitability all working well in small caps. In this episode of Inside Active, host David Cohne, mutual-fund and active-management analyst with Bloomberg Intelligence, along with co-host Christopher Cain, BI’s US quantitative strategist, spoke with Joel Schneider, deputy head of portfolio management at Dimensional Fund Advisors about the firm's systematic, daily and flexible investment process. They also discussed the difference between passive and indexed investing, the hidden costs of indexing and why combining multiple factors can provide better risk control in portfolios.
This podcast was recorded on Jan. 28.
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:14):
Welcome to Inside Active, a podcast about active managers that
goes beyond sound bites and headlines and looks deeper into
their processes, challenges, and philosophies and security selection. I'm David Cohne,
I lead mutual fund and active research at Bloomberg Intelligence.
Speaker 2 (00:29):
Today.
Speaker 1 (00:29):
My cost is Christopher Kine, us quantitative strategist at Bloomberg Intelligence. Chris,
thanks for joining me today.
Speaker 3 (00:36):
Thank you so much for having me.
Speaker 2 (00:37):
David, So, I wanted.
Speaker 1 (00:39):
To ask you about the small cap BMVP portfolio, as
I really think it's partin into our discussion today. How
has the focus on certain factors helped that portfolio up
perform an index like the Russell two thousand.
Speaker 3 (00:53):
Sure, so, our small cap BMVP multi factor portfolio is
very similar to our large cap version. So it's a
only multi factor portfolio using four main factors, which is value, momentum,
low volatility, and profitability. So our small cap version beat
the Russell two thousand equweight index by about ten percentage
(01:14):
points this year or last year, I should say, up
about twenty percent versus about ten to eleven percent for
the index. You know, it's not really a mystery why.
You know, most of the factors did work in small
caps last year similar to large caps. Would I would
flag that value seem to work a bit better, you know,
in small caps, but the devils and the details kind
(01:34):
of about how you define value. But yeah, it's been
a strong factor year for both market capitalizations and hopefully
continues great well.
Speaker 1 (01:43):
I think our guests can add some color to our
discussion on factors. I'd like to welcome Joel Schneider, Joel's
deputy head of portfolio management at Dimensional Fund Advisors. Joel,
thank you for joining us today.
Speaker 2 (01:55):
Hey David, Hey Chris. Happy to be here.
Speaker 1 (01:58):
So before we die into factors and discussion in general,
we'd love to hear how you got your start in
the investment business.
Speaker 2 (02:07):
Yeah. Sure. I'll start with the short answer first, which
is serendipity, meaning I didn't really set out to work
in investment management. The longer answer is, you know, I
was one of those inquisitive kids that used to take
things apart to understand how they worked, and it's probably
no surprise I wound up in engineering, but before I
(02:27):
got there. After school, growing up, I used to go
over to my grandparents' house and my parents both worked,
so my grandparents watched US for a couple hours, and
my grandpa would always watch CNBC, and so it was
riveting for me, as a young person that like numbers,
to see all these prices flying by, to see all
(02:48):
these company fundamentals being talked about. But there was sort
of a randomness to the stock market that never really
made a lot of sense to me, so I sort
of put that interest aside. I ended up going to
college and studying computer engineering, and then I worked at
Lockey Martin and I designed communication systems for the Navy
and the Air Force. And as I was there, I
(03:10):
began to realize that if I was going to progress
in that role, I needed to learn corporate finance. So
to be the head of a large program, or to
be the head of a division, you had to have
your own P and L. So I needed to understand
corporate finance better than I did. So I went back
to business school at University of Chicago to get an MBA,
(03:33):
and while I was there, I learned that there were
a lot of theories and frameworks and evidence that actually
explained how financial markets worked, So to me, this was
really cool. There were valuation frameworks like discounter cash flow
models that help you value companies. There were portfolio theories
that talked about the benefits of diversification based on the
(03:54):
covariance of different assets. There was arbitrage pricing theories that
helped explain how the prices of different financial instruments were
all kept inlign with each other. And then importantly for
today's discussion, there were factor pricing models that helped explain
what drove returns in equities and bonds. And so after
seeing all this, and actually really importantly to me was
(04:18):
these weren't just theories. People were actually rigorously testing them
using the scientific method which I'd come to learn in engineering.
So all of this allowed me to see through the randomness.
And when I graduated from University of Chicago, there's really
no place other that I wanted to work than dimensional
because we'd been associated with a lot of the academics
(04:39):
that founded a lot of those frameworks and did a
lot of that research.
Speaker 1 (04:45):
That's great, and so you know, actually, let's talk about
dimensional or you know DFA as I call it. And
I'm sure a lot of others, and I know many
of our listeners know of DVA, but we'd love to
hear you know, from your I guess experience working there,
you know the basic tenants behind the DFA investment philosophy.
Speaker 2 (05:04):
Yeah, sure, if you don't mind to understand our philosophy,
and might be helpful if I spend a couple minutes
describing our background in our history. So, yeah, we've been
around for about four decades and we've become one of
the largest investment managers and more recently the largest active
ETF manager in the world, and we manage about eight
(05:25):
hundred billion dollars of publicly traded securities, so across equities,
fixed income, real estate, and commodities. And we have this
really deep academic heritage and so a lot of our investments,
and specifically our investment philosophy that you asked about, is
rooted in research that's been done by multiple Nobel Prize
(05:45):
winners as well as many other leading academics. And what
we see as our job at Dimensional is to implement
their research in real world portfolios. So what that ends
up looking like is we run very low cost, broadly
diversified portfolios, but they are not exactly indexed. So we
(06:10):
have active strategies that actually outperform our benchmark index is
at a greater rate and over longer periods of time
than most other managers. In fact, I'm not aware of
anyone that has sort of the types of numbers we have,
So just to give you a sense of what those are,
if you were to go back over the last decade
(06:30):
and say what percentage of managers beat their benchmark index
in the industry, it's a pretty low number. It's only
twenty three percent of managers. But at dimensional seventy eight
percent of our funds have beating their benchmark index over
the last ten years. And if you extend that over
twenty years, it gets even worse. For the industry, only
(06:51):
eighteen percent of funds have beaten their benchmarks, but ninety
two percent of our funds have. And so I want
to get into the investment philosophy now to sort of
explain how we do that. But a key insight, remember
when I said our job is to implement the best
ideas in finance. Well, our co founders actually founded some
of the first index funds in the early nineteen seventies,
(07:14):
and in doing that they came to realize that trading
in a really rigid way where you have to buy
and sell the stocks that the index tells you to
is a recipe for high trade and cost, and so
they realize back then that there's a difference between passive
and indexed. So I think this may help us in
(07:35):
our conversation today. But I just wanted to find passive
and active, or sorry, passive and index For passive, I
think that means just treating market prices is fair and
generally holding securities at their marketcap weight. For indexed, that
is an implementation approach. It's basically where you are forced
(07:58):
to buy and sell securities when some third party index
provider tells you to. And so a lot of the
key to doing better than indexes is avoiding that type
of implementation. All right, So that was a really long
way to get to your question. But in terms of
our investment philosophy, I think it boils down to in
(08:19):
competitive liquid markets, prices are generally fair, they're forward looking,
and they already represent a consensus prediction about the future.
And so the challenge that a lot of traditional active
managers have had is that they're trying to outguess those
market prices. And as we all saw yesterday in the news,
(08:42):
all of a sudden there was news that the Chinese
company behind deep Seek had this great new AI model
that challenged the business model of some of the US
based companies and the suppliers of chips, and so really
quickly prices adjusted, and I didn't really see a lot
of people out there calling that ahead of time. So
(09:04):
markets are moving in real time. Prices are adjusting, and
so our philosophy is embrace that, just make use of it.
And so it starts with just saying the prices are
what they are, they're a reasonable prediction. How do we
bring other pieces of information to combine with the price
(09:24):
to understand which stocks have higher or lower future expector
returns given that price today? And so that to many
people they call that factor investing. But you're combining different
variables or financial metrics from companies' income statements or balance
sheets with the price to make some inferences about which
(09:45):
stocks are likely to do better in the future.
Speaker 1 (09:49):
No, that makes sense. I do want to touch upon
what you said about indexing and passive and so, you know,
this being a podcast focused on active management, if you
could further elaborate on what you consider the inefficiencies of indexing.
You know, you mentioned you know the news yesterday and
you know obviously, you know, holding an index, you know
(10:09):
that that becomes an issue, and so i'd just love
to hear more about that.
Speaker 2 (10:14):
Yeah, for sure. So again, indexing to us is an
implementation decision, and it's basically outsourcing your trading decisions to
some third party index provider. And I think there's multiple
issues when you outsource to them. Now, if we step back,
(10:36):
a lot of people think indexing is really sort of
low cost, and I would actually take issue with that.
I think it's low fee, but it's not necessarily low cost.
And what I mean is that there's a lot of
hidden costs involved with indexing. A big one of those
is that the indexers are all forced to trade the
(10:57):
same names on the same day, the same time as
all the other indexers. And so a really fun analogy
is it's sort of like buying roses on Valentine's Day?
Do you think you're going to get a good price? No, Right,
if you bought roses a week or two before or after,
you're going to get a much better price than if
you're buying them on that day. So that's called the
index reconstitution effect. And you guys may know these numbers
(11:21):
better than me, but if you look at the growth
of indexing, the total dollars chasing after the same names,
I saw numbers las year at the end of last
year around like twelve trillion dollars in index products. So
that means that when there's these rebalance events, you've got
billions of dollars, tens of billions of dollars that are
(11:42):
all chasing the same stocks. And what that does is
it tends to on average, push up prices of names
that are being added to the index and pushed down
prices of names that are being dropped. And our research
we actually just did an updated set of research papers
on this. There's a lot of academic work about ten
years ago, so we decided to do an update for
(12:04):
the next decade. We looked at both US and non
US equity indices, and we found that prices get pushed
by about four percent on average. Four percent. I don't
know if that feels like a big or small number
to the people listening, but let me just put that
in context a little. Let's say an index has five
(12:25):
percent turnover a year. It's pretty low turnover. Eight Well,
if five percent of your portfolio is getting four percent
worse prices that's a potential drag of twenty basis points
a year. Now, when people think of the low expense
ratios or management fees of indexing, that twenty basis points
of hidden performance drag is in many cases much bigger
(12:48):
than the fee they're paying, So really their total cost
of ownership is a lot higher than they think it is.
Speaker 3 (12:53):
Now.
Speaker 2 (12:54):
The challenge with seeing that is both the index and
the index fund suffer from that because they're both adding
the stocks at the close on the rebalanced day. So
your index fund will have maybe no tracking here with
your index, but both of them have that performance drag
(13:15):
baked in. So I think that's one of the biggest
inefficiencies of indexing. I think there's other ones, though we
can get into this a little bit later, especially when
you're trying to capture factor. Premium indexes have a lot
of style drift, and that becomes a big issue. I
heard Chris at the beginning talking about capturing some of
the premiums within small cap like profitability and value. Well,
(13:39):
if the Russell two thousand is only rebalancing once a
year in June, then that means they've got eleven months
where the stocks they hold have drifted. Many have become
midcaps or large caps, and so the index style drift
is a major issue. And then you know, I talk
with a lot of institutional clients and one of the
(14:00):
things that they've come to really realize is that a
lot of their index managers may be charging them a
very low management fee, but then they're keeping a pretty
large percentage of the securities lending profits from those stocks,
and in some ways that serves as sort of a
shadow management fee, right it looks like you're only paying
(14:20):
a couple BIPs and expensory sue or in management fee,
but then they're keeping a slug of the sect lending
revenues for themselves. So those are just some of the issues.
I'm sure we'll get into more of them later. But
there's there's definitely issues with index implementation, and you can
do better than indexing by not being so rigid.
Speaker 1 (14:39):
No makes sense, and so you know, switching from passive
over to active, and you know, I know Chris has
a bunch of questions for you in terms of you know,
factors as you touched upon a little bit, but I
would just love to hear you know, what factors in
the research at Dimensional have you found that historically driven performance.
Speaker 2 (14:59):
Yeah, I think the answer to that depends on what
time frame you're measuring. So we like to think about
three different timeframes. So in the long term, let's say
we're measuring over a year or more. In terms of
those long term drivers, it's company's valuation, their profitability, and
(15:23):
their size. Then you start to get into some of
the more intermediate or shorter term drivers. Let's say we're
measuring those over months or weeks. They're things like momentum
or asset growth, or interestingly, stocks that are expensive to borrow.
(15:44):
In the securities lending market, that's actually, to me a
really cool factor. We often say that we extract information
from market prices. So earlier you asked me about our
investment philosophy, and I said, we take the prices for
what they are and we see what information we can
extract from that. Well, here's an example where the securities
(16:05):
lending market that is another market just like the stock
markets a market, the stock lending markets market, and it's
got prices, and if people are willing to pay very
high fees to borrow stocks, often to shorten them, that's
actually a really negative sign that explains under performance. Of
stocks over the next few weeks after they become expensive
(16:26):
to borrow. And then the last very sort of short
term factors are things related to liquidity. So one of
them that's in the academic research is price reversals, and
then the other are things related to implicit trade and costs,
so spreads, price impact. So I think, really zooming back out,
(16:52):
you want to think about which factors are reliable, but
then which ones apply over different timeframes, and how you
implement will depend on the timeframe that they apply over.
And so I think maybe the last thing to add
before I'm sure you have questions on that is some
listeners may be wondering why I didn't mention some of
(17:14):
the commonly cided factors that they've heard of. And unfortunately,
you guys have probably heard the term the factor zoo.
So there's so many people publishing papers about different factors.
You know, some people say three hundred or four hundred
factors have been identified. Well, those are mainly filled with
either deplicative or unreliable factors. And so when I say deplicative,
(17:37):
I just mean that once you control for the factors
I already mentioned that the other ones don't add any
new explanatory power. It's not to say there's anything wrong
with those, it's just it could be you know that
you're going to approach that factor using some other definition.
That's fine as long as you know and you're not
being redundant and applying basically sort of the same factor twice.
(18:01):
And then the other reason why I didn't mention some
is some of them just don't pass the high standards
of the scientific method which I talked about earlier, and
that is you need to have a strong economic theory
for why a factor should exist. You need to be
able to reproduce those results, and they need to be
able to hold up an out of sample testing. Otherwise
(18:23):
it's hard to be confident that those will occur in
the future.
Speaker 3 (18:27):
So interesting, I you know, it's a jewel. I reserve
the right to steal your buying roses on Valentine's Day analogy.
I love that. So my questions around combining multiple factors,
I mean, you know, thank you for walking through that.
That was really interesting. With the different timeframes, you know,
I don't see many people frame it that way, So
(18:48):
you know, can you talk us through how those different
timeframes you know, apply to a multi factor process. How
do you combine factors, especially if they have different time frames?
It is any element, and I guess, I guess I
can ask do you believe in factor timing, Like, do
you think there's a way to time factors or do
(19:08):
you think it's a better approach to just have a
relatively constant exposure to factors that you believe, you know
are are advantages for the long term.
Speaker 2 (19:21):
Sure, well, it's a two part question, So let me
take the second part of your question first, which is
we have looked for every way that we could possibly
think of time factors, and we really wish that you
could just Unfortunately there's no evidence that you can, and
that oftentimes the cost of getting it wrong is really significant.
Speaker 3 (19:44):
Right.
Speaker 2 (19:44):
One of the things I learned in engineering is you
always have to think about if a certain part or
system fails, how bad is the outcome when it fails?
And getting factor timing wrong can be a really expensive
and so generally it's better to take multiple factors that
(20:06):
are reliable but that oftentimes are not highly correlated with
each other and include them into a multi factor portfolio.
This is going to give you a little bit more
risk control and the ability to sort of ride out
or survive different periods in the market when certain factors
are in or out of favor. So I think, Chris,
(20:28):
that gets to your second question, which is, then how
do you start to combine multiple factors into a portfolio?
And I think there's a few lessons to keep in mind.
The first one that I said earlier is, you know,
more factors are not necessarily better. There's a quote that's
often attributed to Einstein that I really like that you know,
(20:51):
supposedly pretty much every quote is either attributed nowadays to
either Einstein or Mark Twain, so you never really know
if they're said, but anyway, it's a good quote which
says everything should be as simple as it can be,
but not simpler, right, So there is room for things
that are complicated in this world. But just throwing additional
(21:14):
factors into something, even though it may seem sophisticated, sometimes
actually is detrimental. And so I would say you want
to start with factors that are rigorously tested or not
duplicative with one another, and then, like you said, understand
the timeframes. And so the way that we approach it
is those long term factors that help explain returns over
(21:37):
years those are good things to actually build a strategy's
construction around, right, because you can do that in a
pretty stable way without a lot of turnover, and so
things like you mentioned earlier you were talking about that
small cap strategy, So things like valuation, profitability, size, those
(21:57):
are great to include in a long term strategy. And
the way that we do it is we will start
with market cap weights of securities, and then to the
extent that securities look good across multiple factors are bad,
we'll overweight and underweight relative to market cap weights. Now,
then this brings in some of those shorter time period factors,
(22:21):
and so with those, I think it's important to not
apply them in the construction because if a factor is
changing its signal or it's information is providing you every
couple of weeks, then it's going to cause a lot
of turnover in the portfolio. That could cause very high
trading cost or if you have taxbile investors, that could
(22:42):
cause them a huge capital gains tax bill. So there,
I think the best thing to do is to use
them as delays. So let's say that you would have
purchased or sold some security based on the long term factors,
but then you screen them for the short term factors,
and then you may decide to either delay that buy
(23:04):
or sell and then substitute in another name that isn't
having that maybe negative short run expected return. And so
I think that tends to be the way that we
think about it. And then the last thing that I
would say is you really want to understand how those
factors interrelate with one another. And so for example, value
(23:26):
and profitability, they tend to be complements. So more often
than not, when the value premium is negative, the profitability
premium is positive, or vice versa. So therefore those are
great to combine in a portfolio, whereas profitability and let's
say growth, those tend to be positively correlated. So if
you aren't careful, you could just be doubling down and
(23:49):
increasing your risk without really increasing your expected return. So
I think that's the main way that we think about
including multiple factors.
Speaker 3 (23:57):
That's so interesting. Thank you. I mean, I think it's
relative unique that you guys do like the you know,
the long term and in the short term, and I
thank you for explaining that. I mean, that's really a
really cool perspective. Who knows if Leonardo da Vinci actually
said this, But people say, Leonordo da Vinci said, simplicity
is the ultimate sophistication. That's what that's I like that.
Speaker 2 (24:18):
Yeah, that's great. So you can borrow my Roses's Day
and I'll borrow your supposedly Anaro DaVinci.
Speaker 3 (24:24):
And who knows if you said it or not. But
you know what, when you when you name drop Leonardo
da Vinci, you sounds smart. So there you go.
Speaker 2 (24:31):
Maybe it was Mark Twain, Yeah.
Speaker 3 (24:32):
Exactly, who knows it was Yogi Berra? No? Yeah, you
kind of let me do another question I had so,
you know, like I write a lot about factor investing
and and sometimes people that maybe you know, certainly aren't
as sophisticated as you and might not know much about
this stuff, they'll come back to me and say, why
(24:53):
don't you have growth as a factor? Is in growth
a factor? You know? What's the difference between something like
profitability or maybe a more broad definition you would say
quality and value? I'm sorry, and growth? Are they the
same thing? Is one better than the other? Why do
(25:13):
you always talk about quality slash profitability and our growth?
What would you say to a question like that?
Speaker 2 (25:19):
Yeah, I'm glad you bring it up because I do
think it's confusing for a lot of people, and they
often conflate these different factors, and so I see the
same thing when I talk with people, and I think
some of it just comes down to not being clear
about how these things are defined. And so for us,
(25:41):
we define value as companies that have low valuations. So
you can use various metrics. The good news is that
they all actually contain some information. But some companies have
low valuations, some have high. To us, the low valuation
or low relative price is value and the high relative
price is growth. Where when you get into quality and look,
(26:05):
I know other people have different definitions of growth, including
companies that are growing their earnings, which is also kind
of related to profitability and momentum. There's some interesting work
Robert Novi Marx who's a professor at the University of Rochester.
He's looked at momentum and profitability growth, and so there's
sort of a version of earnings momentum that's really interesting research.
(26:30):
But at least those you can tie them to specific
line items on a company's income statement or balance sheet.
I think with quality, unfortunately, I have to say it's
a bit more of a marketing term. Than a financial term.
And what I mean is I think it's designed to
appeal to people's sort of intuitive sense of, oh, well,
(26:53):
this company has quality earnings or quality balance sheet, but
there's no standard definition of that, and so I think
this causes a lot of the confusion. And when we
look at the research on quality, there's a number of
variables that managers tend to use, so return on equity, leverage, earnings, variability, others.
(27:21):
And unfortunately, when you add those factors in to a
model that already contains profitability, they don't add any additional
explanatory power. So I think for listeners, if you have
a portfolio that is already focused on valuations, profitability, and
(27:41):
then momentum considers momentum as well, you're pretty much picking
up all the effect that you're going to get from
both quality and growth.
Speaker 3 (27:51):
I couldn't agree with that more. I mean, even in
my own work, you know, I've kind of moved away
from saying quality for those exact reasons you said. People
have different ideas, and you know it does play on
people's like, of course you want high quality on low quality, right,
who wouldn't. Yeah, But I mean in my research and
I'm sure you'd agree with this, Like, profitability is by
far the biggest driver of quality, and the other things
(28:13):
really add negligible value, and so why don't we just
use profitability. It's much more easy to understand, and I
think that the you know, the research there is kind
of more clear. So I totally agree with you.
Speaker 2 (28:24):
Yeah, I agree with you, and I think Leonardo da
Vinci would agree with you as well.
Speaker 1 (28:29):
Talking about factors, you know, you mentioned long term factors
kind of you know, it's the basis for the portfolio management.
So certainly if you can kind of go into the
investment process a little bit of you know, from a
manager's standpoint of, you know, what are they? What is
the process of, you know, taking this research and implement
that into an actual portfolio or you know, our funds.
Speaker 2 (28:53):
Yeah, sure, I think I would describe our investment process
using three words. The first is stematic, the second is daily,
and the third is flexible. So let me say what
I mean by that. Every single day we take current
market prices and the most recent company fundamentals and we
(29:17):
use that to assign companies to in terms of different factors,
so value or profitability. Also, we then calculate sort of
theoretical weights for every security in every portfolio. And I'm
really gonna highlight the word theoretical there because this is
only using the long term factors, so we haven't enriched
(29:39):
that with additional information about the short term ones yet.
So it's sort of a rough work and process, if
you will. And so every day, though, we have a
description of those securities and where they sit across the
different factors, and that's very different from an indexed based approach.
(30:00):
So let me just contrast it real quick. If you're
invested in a value index, for example, or a profitability
or quality or whatever index, they're only doing those updated
calculations and bucketing of securities maybe at most every quarter. Oftentimes,
even when they say they have quarterly rebalances, they're only
(30:22):
really redefining the breakpoints between those factors on a semi
annual basis, So they're working with stale information. And so
we have a daily process to combine securities and financial
metrics categorize stocks. At that point, our portfolio managers will
(30:46):
review that updated information and where those stocks are sitting,
and we'll compare it to our current holdings, and so
then that may suggest that we may want to do
some rebalancing. So you're familiar that in X is rebalanced
maybe a couple times a year. Well, we rebalance a
little bit every single day, so it's like having two
(31:06):
hundred and fifty or so rebalance events throughout the year.
And so every day we're looking at the cash flows
coming in and out of the portfolio and thinking, how
do I use those as efficiently as possible. If there's
some security that became a lower valuation or more profitable,
and we want to increase our weight in that, how
do we use the cash flows coming in from either
(31:28):
clients or maybe from corporate actions like dividends, how do
we redeploy that cash just to the stocks that we
want to increase our weight in. And then that's when
we start to apply those shorter term criteria, those shorter
term factors, And what that will do is it will
cause us to delay from trading some of the names.
(31:50):
So we will say, all right, well, there's maybe ten
securities in the US arch cap space that we want
to buy more of, but three of them maybe have
these short term negative factors, We'll go by the other
seven instead. And from there what we do, and this
is this part becomes very unique now compared with anyone
(32:11):
else to know in the industry, is we will send
those over to our traders, and let's just say hypothetically
that we want to spend fifty million dollars, we will
give our traders, let's say three times that amount of
order candidates, So we'll give them one hundred and fifty
million in order candidates, and we will give them the
(32:31):
exact share counts and price limits and everything, so they
don't have discretion on which securities to eventually buy. But
what we do is we give them flexibility over the
timing and the quantity. So we say, buy anything off
of this list today, We'll come back and do it
tomorrow and the next day and the next day. And
so what ends up happening is our portfolio managers get
(32:54):
all the positions they want, but our traders also get
the flexibility to not have to cross spreads or push prices.
And so this helps avoid an issue that both index
funds and active managers have, which is, in some ways
most of them are demanding liquidity from the market. They're
(33:15):
going and saying I need to trade a specific stock
in a specific quantity at a specific time and when
you do that, you just don't get great prices. Whereas
if our traders can sit over on the favorable side
of the spread and let other people cross and we
can get some price improvement, that actually is a value
add in our process. So that's how the process works.
When we say it's systematic, we've built systems to do
(33:40):
this daily rebalancing in the lowest cost way that we can.
Speaker 3 (33:46):
Really interesting, it's like buying one ros a day going
up to Valentine's Day.
Speaker 2 (33:53):
Yeah, Chris, I've tried to extend this analogy where sometimes
I say, we give our traders a shopping list and
tell them to go to the grocery store, but then
we only give them a budget to buy like a
third of the shopping list, and so every day they
have to buy what's on sale. It's just at some
point you stretch the analogy so far that people are like,
who would go to the grocery store that often? But
(34:14):
with electronic trading, actually you can go to the grocery
store all the time. It's fine.
Speaker 3 (34:18):
Sure, Yeah, And I bet that working those orders, I
bet that means you know, a big difference over time.
I could totally see that, and like you said, I mean,
when you have an index, everyone knows when you're rebalancing,
and it's not you know, you can front run that stuff.
All right, let me ask you a question. This is
like as controversial as it gets, right when it comes
to factors small size. So when I first learned factor investing,
(34:42):
it was like small size as a factor. And by
a factor, I don't mean like a risk factor, I
mean like an alpha factor, like it's going to give
you higher risk adjusted returns. And then I feel like
over the last decade or two, the evidence of small
size being an actual premium has really been hit, and
some argue that it was never a premium at all.
(35:03):
It was just you're taking tail risk because you're buying
small companies and they could go bankrupt and there's more
volatility and more left tail risk there, so you should
be compensated for higher returns. It's not actually a premium.
As you know, small sizes in many you know academic
factor models. So where do you come down on this debate?
I mean, do you think small size is still a
(35:24):
factor or have you kind of reassess that over the
last couple of years.
Speaker 2 (35:29):
Yeah, So I think two things. The first is it's
always important to continuously reassess the evidence. No one should
ever just stick with what was done at some point
in the past. Being a statistics nerd myself, for those
other listeners out there that are this is called taking
(35:50):
a basin approach. It's you have your priors, you get
new data, you update you know, you add it to
the information, you update your priors. And so that's a
big part of what our research team does here is
continuously test if these things are still reliable. And then
I think the other thing to point out is in
the US we have seen a couple decades where small
(36:13):
caps have done worse than large caps, which I think
is interesting for two reasons. One is that hasn't been
the case outside the US, so across all the other
non US countries in aggregate, we've actually seen positive size premiums.
And then in the US, I think there's two things
going on. One is that people tend to use the
(36:36):
Russell two thousand as their proxy for how small caps
have done. But we talked earlier about some of the
performance drag associated with index reconstitution, and so we said
that's coming out of the index. So really that small
cap index has particularly low returns compared with some other
(36:56):
small cap indices. And then the other thing else is
you asked me earlier about multiple factors. You always have
to control for all the other So size is just
a one dimensional concept. Actually, isn't that helpful? Right? It's
kind of like if you were a medical researcher and
you said, like, did someone eat a healthy diet. Well,
(37:22):
I'm a cyclist, and so I actually eat a lot
of sugar, but I need that for exercise, and the
net effect is it's good for me in some quantity
at some time. So you always have to think about
multiple explanatory variables. And with small the issue in the
US actually has not been most small cap stocks. It's
been a very small subset of small cap stocks that
(37:45):
have very high valuations and very low or often negative profitability.
And so some people like to sort of casually say,
you know, there's a lot of junk and small caps,
and so I think in small cap investing you have
to be very careful that you are controlling for those
other factors as well. Once you do that, yeah, there
(38:07):
is a premium for small cap stocks, and we tend
to see that some of the other premiums are actually
a little bit stronger in small cap than in large.
Speaker 3 (38:15):
Yeah, I found that too, of the other factors, and
you know, just simply doing a profitability screen on the
Russell two thousand goes a long way, you know, like.
Speaker 2 (38:24):
Absolutely, yeah, absolutely, I mean it's sort of we like
to geek out about factors and be quantitative, but just
step back common sense. Wise investing is always about what
am I paying for something versus what am I expecting
to get right, And if you pay a high valuation
(38:45):
for something that has little to no profits, that's not
generally regarded as a high expected return investment. And so
I think it's always important in any area of the
market to use that framework, but especially in small caps
because there are just a lot of those unprofitable small
cap companies in the US.
Speaker 3 (39:05):
All right, let me ask you about kind of the
topic of the day, which is AI machine learning. You know,
you you mentioned deep seek. That was the you know,
big topic of the markets yesterday. Have you found you know,
applications of it could be just mL models or even LM,
you know, AI models in factor investing or if you know,
(39:28):
if not, like, do you think that's a thing we're
going to you know, really focus on in the future
trying to have machine learning and such help us with
the factors. What's your thoughts there.
Speaker 2 (39:40):
Yeah, it's something we've looked into a lot. And the
interesting thing you mentioned deep Seek again is they actually
started out as a quantitative hedge fund, and that hedge
fund ran into some troubles and so they've sort of
pivoted over to a generative AI model. And I think
it just highlights that it's been very difficult to apply
(40:02):
AI to try to whether it's time the market or
find new factors. In some ways, you start to worry
about what researchers would call data mining, which is you know,
we talked about the four hundred factors and the factor zoo. Well,
you could probably enumerate hundreds of thousands or millions of
(40:23):
combinations of different variables if you were to combine things
off of the income statement and the balance sheet, so
you would make all sorts of esoteric variables and then
you would run them all through a factor model and
by chance some would work right. So this is called
pa hacking in the statistics community. It's one of those
things where you know, even if there's only a five
(40:45):
percent chance that something happens by random. If you run
one hundred tests, well, then five of them are going
to turn up positive, even though they're just by chance.
So i'd mentioned Robert nobe Marx. He actually wrote a
paper on this topic recently. He not only only created
a bunch of AI models to find new factors, he
also then employed AI models to write the academic papers
(41:08):
for him. And it was done as sort of a
way to demonstrate to the industry what can go wrong
if you use this approach. So it is something we've
looked at a lot so far. The ways that we
have experimented with using it is more to gather some
of that unstructured data that may be like in company
(41:29):
filings so or oftentimes when companies have corporate actions, they're
putting out a bunch of text based information. And if
you're using different factors, so for example, if you're using
a profitability factor a value factor, it's relying on the
numbers in the income statement to be interpreted properly. And
so let's say that you know the auditor qualifies their
(41:53):
opinion about the income statements, Well, do you want to
trust the metrics from there? Probably not. So if you
had some sort of a model that could scan through
and highlight to you areas where there may be concerns
or maybe changes. So let's say a company spins off
a division. Well, now anything on the balance sheet about
you know, assets or book value of equity has changed
(42:15):
because they've spun some of that off. So finding ways
to flag some of those changes in the data so
you can go update your metrics so you're not using
stale metrics has been an area where we've experimented with
using it.
Speaker 1 (42:29):
Oh, this is great. We just have one more question
before we let you go. You know, like to ask
this a lot to a lot of our guests. But whatever,
you're some of your favorite financial or investing books.
Speaker 2 (42:40):
Yeah, so I read financial literature all day long at
my job, and so when I get a chance to
read books, I tend to like books that come from
other fields. And I really enjoy history. I think we
have a lot to learn from history, and specifically I
like biographies, and so one of my favorite authors is
(43:02):
Walter Isaacson, and he's written biographies on people like Steve Jobs,
Leonardo da Vinci, which you were talking about earlier, Elon Musk,
Benjamin Franklin, and others. And one of the things I
really like reading about in these biographies is each of
these individuals accomplished a lot. They were all great problem solvers,
(43:27):
and each of them went through a lot of challenges.
They went through big ups and downs along the way.
And I think intellectually we all know that success, whether
that's in life or in investing, it's not a line
that goes straight up into the right right. But there's
a difference between knowing that and then actually being able
(43:48):
to have the discipline to sort of live through the
downs and capture the ups. And so for me, reading
biographies really brings to life the lessons from history, so
I can learn from them and not have to repeat
them myself. It's great, Joel.
Speaker 1 (44:04):
I enjoyed this. Thanks again for joining us today.
Speaker 2 (44:06):
Absolutely, Thank you for having me and Chris.
Speaker 1 (44:08):
Thank you for being my co host again today.
Speaker 3 (44:11):
Thank you, and thank you Joel.
Speaker 1 (44:12):
Until our next episode, this is David Cohne with the
Inside Out.
Welcome to Inside Active, a podcast about active managers that
goes beyond sound bites and headlines and looks deeper into
their processes, challenges, and philosophies and security selection. I'm David Cohne,
I lead mutual fund and active research at Bloomberg Intelligence.
Speaker 2 (00:29):
Today.
Speaker 1 (00:29):
My cost is Christopher Kine, us quantitative strategist at Bloomberg Intelligence. Chris,
thanks for joining me today.
Speaker 3 (00:36):
Thank you so much for having me.
Speaker 2 (00:37):
David, So, I wanted.
Speaker 1 (00:39):
To ask you about the small cap BMVP portfolio, as
I really think it's partin into our discussion today. How
has the focus on certain factors helped that portfolio up
perform an index like the Russell two thousand.
Speaker 3 (00:53):
Sure, so, our small cap BMVP multi factor portfolio is
very similar to our large cap version. So it's a
only multi factor portfolio using four main factors, which is value, momentum,
low volatility, and profitability. So our small cap version beat
the Russell two thousand equweight index by about ten percentage
(01:14):
points this year or last year, I should say, up
about twenty percent versus about ten to eleven percent for
the index. You know, it's not really a mystery why.
You know, most of the factors did work in small
caps last year similar to large caps. Would I would
flag that value seem to work a bit better, you know,
in small caps, but the devils and the details kind
(01:34):
of about how you define value. But yeah, it's been
a strong factor year for both market capitalizations and hopefully
continues great well.
Speaker 1 (01:43):
I think our guests can add some color to our
discussion on factors. I'd like to welcome Joel Schneider, Joel's
deputy head of portfolio management at Dimensional Fund Advisors. Joel,
thank you for joining us today.
Speaker 2 (01:55):
Hey David, Hey Chris. Happy to be here.
Speaker 1 (01:58):
So before we die into factors and discussion in general,
we'd love to hear how you got your start in
the investment business.
Speaker 2 (02:07):
Yeah. Sure. I'll start with the short answer first, which
is serendipity, meaning I didn't really set out to work
in investment management. The longer answer is, you know, I
was one of those inquisitive kids that used to take
things apart to understand how they worked, and it's probably
no surprise I wound up in engineering, but before I
(02:27):
got there. After school, growing up, I used to go
over to my grandparents' house and my parents both worked,
so my grandparents watched US for a couple hours, and
my grandpa would always watch CNBC, and so it was
riveting for me, as a young person that like numbers,
to see all these prices flying by, to see all
(02:48):
these company fundamentals being talked about. But there was sort
of a randomness to the stock market that never really
made a lot of sense to me, so I sort
of put that interest aside. I ended up going to
college and studying computer engineering, and then I worked at
Lockey Martin and I designed communication systems for the Navy
and the Air Force. And as I was there, I
(03:10):
began to realize that if I was going to progress
in that role, I needed to learn corporate finance. So
to be the head of a large program, or to
be the head of a division, you had to have
your own P and L. So I needed to understand
corporate finance better than I did. So I went back
to business school at University of Chicago to get an MBA,
(03:33):
and while I was there, I learned that there were
a lot of theories and frameworks and evidence that actually
explained how financial markets worked, So to me, this was
really cool. There were valuation frameworks like discounter cash flow
models that help you value companies. There were portfolio theories
that talked about the benefits of diversification based on the
(03:54):
covariance of different assets. There was arbitrage pricing theories that
helped explain how the prices of different financial instruments were
all kept inlign with each other. And then importantly for
today's discussion, there were factor pricing models that helped explain
what drove returns in equities and bonds. And so after
seeing all this, and actually really importantly to me was
(04:18):
these weren't just theories. People were actually rigorously testing them
using the scientific method which I'd come to learn in engineering.
So all of this allowed me to see through the randomness.
And when I graduated from University of Chicago, there's really
no place other that I wanted to work than dimensional
because we'd been associated with a lot of the academics
(04:39):
that founded a lot of those frameworks and did a
lot of that research.
Speaker 1 (04:45):
That's great, and so you know, actually, let's talk about
dimensional or you know DFA as I call it. And
I'm sure a lot of others, and I know many
of our listeners know of DVA, but we'd love to
hear you know, from your I guess experience working there,
you know the basic tenants behind the DFA investment philosophy.
Speaker 2 (05:04):
Yeah, sure, if you don't mind to understand our philosophy,
and might be helpful if I spend a couple minutes
describing our background in our history. So, yeah, we've been
around for about four decades and we've become one of
the largest investment managers and more recently the largest active
ETF manager in the world, and we manage about eight
(05:25):
hundred billion dollars of publicly traded securities, so across equities,
fixed income, real estate, and commodities. And we have this
really deep academic heritage and so a lot of our investments,
and specifically our investment philosophy that you asked about, is
rooted in research that's been done by multiple Nobel Prize
(05:45):
winners as well as many other leading academics. And what
we see as our job at Dimensional is to implement
their research in real world portfolios. So what that ends
up looking like is we run very low cost, broadly
diversified portfolios, but they are not exactly indexed. So we
(06:10):
have active strategies that actually outperform our benchmark index is
at a greater rate and over longer periods of time
than most other managers. In fact, I'm not aware of
anyone that has sort of the types of numbers we have,
So just to give you a sense of what those are,
if you were to go back over the last decade
(06:30):
and say what percentage of managers beat their benchmark index
in the industry, it's a pretty low number. It's only
twenty three percent of managers. But at dimensional seventy eight
percent of our funds have beating their benchmark index over
the last ten years. And if you extend that over
twenty years, it gets even worse. For the industry, only
(06:51):
eighteen percent of funds have beaten their benchmarks, but ninety
two percent of our funds have. And so I want
to get into the investment philosophy now to sort of
explain how we do that. But a key insight, remember
when I said our job is to implement the best
ideas in finance. Well, our co founders actually founded some
of the first index funds in the early nineteen seventies,
(07:14):
and in doing that they came to realize that trading
in a really rigid way where you have to buy
and sell the stocks that the index tells you to
is a recipe for high trade and cost, and so
they realize back then that there's a difference between passive
and indexed. So I think this may help us in
(07:35):
our conversation today. But I just wanted to find passive
and active, or sorry, passive and index For passive, I
think that means just treating market prices is fair and
generally holding securities at their marketcap weight. For indexed, that
is an implementation approach. It's basically where you are forced
(07:58):
to buy and sell securities when some third party index
provider tells you to. And so a lot of the
key to doing better than indexes is avoiding that type
of implementation. All right, So that was a really long
way to get to your question. But in terms of
our investment philosophy, I think it boils down to in
(08:19):
competitive liquid markets, prices are generally fair, they're forward looking,
and they already represent a consensus prediction about the future.
And so the challenge that a lot of traditional active
managers have had is that they're trying to outguess those
market prices. And as we all saw yesterday in the news,
(08:42):
all of a sudden there was news that the Chinese
company behind deep Seek had this great new AI model
that challenged the business model of some of the US
based companies and the suppliers of chips, and so really
quickly prices adjusted, and I didn't really see a lot
of people out there calling that ahead of time. So
(09:04):
markets are moving in real time. Prices are adjusting, and
so our philosophy is embrace that, just make use of it.
And so it starts with just saying the prices are
what they are, they're a reasonable prediction. How do we
bring other pieces of information to combine with the price
(09:24):
to understand which stocks have higher or lower future expector
returns given that price today? And so that to many
people they call that factor investing. But you're combining different
variables or financial metrics from companies' income statements or balance
sheets with the price to make some inferences about which
(09:45):
stocks are likely to do better in the future.
Speaker 1 (09:49):
No, that makes sense. I do want to touch upon
what you said about indexing and passive and so, you know,
this being a podcast focused on active management, if you
could further elaborate on what you consider the inefficiencies of indexing.
You know, you mentioned you know the news yesterday and
you know obviously, you know, holding an index, you know
(10:09):
that that becomes an issue, and so i'd just love
to hear more about that.
Speaker 2 (10:14):
Yeah, for sure. So again, indexing to us is an
implementation decision, and it's basically outsourcing your trading decisions to
some third party index provider. And I think there's multiple
issues when you outsource to them. Now, if we step back,
(10:36):
a lot of people think indexing is really sort of
low cost, and I would actually take issue with that.
I think it's low fee, but it's not necessarily low cost.
And what I mean is that there's a lot of
hidden costs involved with indexing. A big one of those
is that the indexers are all forced to trade the
(10:57):
same names on the same day, the same time as
all the other indexers. And so a really fun analogy
is it's sort of like buying roses on Valentine's Day?
Do you think you're going to get a good price? No, Right,
if you bought roses a week or two before or after,
you're going to get a much better price than if
you're buying them on that day. So that's called the
index reconstitution effect. And you guys may know these numbers
(11:21):
better than me, but if you look at the growth
of indexing, the total dollars chasing after the same names,
I saw numbers las year at the end of last
year around like twelve trillion dollars in index products. So
that means that when there's these rebalance events, you've got
billions of dollars, tens of billions of dollars that are
(11:42):
all chasing the same stocks. And what that does is
it tends to on average, push up prices of names
that are being added to the index and pushed down
prices of names that are being dropped. And our research
we actually just did an updated set of research papers
on this. There's a lot of academic work about ten
years ago, so we decided to do an update for
(12:04):
the next decade. We looked at both US and non
US equity indices, and we found that prices get pushed
by about four percent on average. Four percent. I don't
know if that feels like a big or small number
to the people listening, but let me just put that
in context a little. Let's say an index has five
(12:25):
percent turnover a year. It's pretty low turnover. Eight Well,
if five percent of your portfolio is getting four percent
worse prices that's a potential drag of twenty basis points
a year. Now, when people think of the low expense
ratios or management fees of indexing, that twenty basis points
of hidden performance drag is in many cases much bigger
(12:48):
than the fee they're paying, So really their total cost
of ownership is a lot higher than they think it is.
Speaker 3 (12:53):
Now.
Speaker 2 (12:54):
The challenge with seeing that is both the index and
the index fund suffer from that because they're both adding
the stocks at the close on the rebalanced day. So
your index fund will have maybe no tracking here with
your index, but both of them have that performance drag
(13:15):
baked in. So I think that's one of the biggest
inefficiencies of indexing. I think there's other ones, though we
can get into this a little bit later, especially when
you're trying to capture factor. Premium indexes have a lot
of style drift, and that becomes a big issue. I
heard Chris at the beginning talking about capturing some of
the premiums within small cap like profitability and value. Well,
(13:39):
if the Russell two thousand is only rebalancing once a
year in June, then that means they've got eleven months
where the stocks they hold have drifted. Many have become
midcaps or large caps, and so the index style drift
is a major issue. And then you know, I talk
with a lot of institutional clients and one of the
(14:00):
things that they've come to really realize is that a
lot of their index managers may be charging them a
very low management fee, but then they're keeping a pretty
large percentage of the securities lending profits from those stocks,
and in some ways that serves as sort of a
shadow management fee, right it looks like you're only paying
(14:20):
a couple BIPs and expensory sue or in management fee,
but then they're keeping a slug of the sect lending
revenues for themselves. So those are just some of the issues.
I'm sure we'll get into more of them later. But
there's there's definitely issues with index implementation, and you can
do better than indexing by not being so rigid.
Speaker 1 (14:39):
No makes sense, and so you know, switching from passive
over to active, and you know, I know Chris has
a bunch of questions for you in terms of you know,
factors as you touched upon a little bit, but I
would just love to hear you know, what factors in
the research at Dimensional have you found that historically driven performance.
Speaker 2 (14:59):
Yeah, I think the answer to that depends on what
time frame you're measuring. So we like to think about
three different timeframes. So in the long term, let's say
we're measuring over a year or more. In terms of
those long term drivers, it's company's valuation, their profitability, and
(15:23):
their size. Then you start to get into some of
the more intermediate or shorter term drivers. Let's say we're
measuring those over months or weeks. They're things like momentum
or asset growth, or interestingly, stocks that are expensive to borrow.
(15:44):
In the securities lending market, that's actually, to me a
really cool factor. We often say that we extract information
from market prices. So earlier you asked me about our
investment philosophy, and I said, we take the prices for
what they are and we see what information we can
extract from that. Well, here's an example where the securities
(16:05):
lending market that is another market just like the stock
markets a market, the stock lending markets market, and it's
got prices, and if people are willing to pay very
high fees to borrow stocks, often to shorten them, that's
actually a really negative sign that explains under performance. Of
stocks over the next few weeks after they become expensive
(16:26):
to borrow. And then the last very sort of short
term factors are things related to liquidity. So one of
them that's in the academic research is price reversals, and
then the other are things related to implicit trade and costs,
so spreads, price impact. So I think, really zooming back out,
(16:52):
you want to think about which factors are reliable, but
then which ones apply over different timeframes, and how you
implement will depend on the timeframe that they apply over.
And so I think maybe the last thing to add
before I'm sure you have questions on that is some
listeners may be wondering why I didn't mention some of
(17:14):
the commonly cided factors that they've heard of. And unfortunately,
you guys have probably heard the term the factor zoo.
So there's so many people publishing papers about different factors.
You know, some people say three hundred or four hundred
factors have been identified. Well, those are mainly filled with
either deplicative or unreliable factors. And so when I say deplicative,
(17:37):
I just mean that once you control for the factors
I already mentioned that the other ones don't add any
new explanatory power. It's not to say there's anything wrong
with those, it's just it could be you know that
you're going to approach that factor using some other definition.
That's fine as long as you know and you're not
being redundant and applying basically sort of the same factor twice.
(18:01):
And then the other reason why I didn't mention some
is some of them just don't pass the high standards
of the scientific method which I talked about earlier, and
that is you need to have a strong economic theory
for why a factor should exist. You need to be
able to reproduce those results, and they need to be
able to hold up an out of sample testing. Otherwise
(18:23):
it's hard to be confident that those will occur in
the future.
Speaker 3 (18:27):
So interesting, I you know, it's a jewel. I reserve
the right to steal your buying roses on Valentine's Day analogy.
I love that. So my questions around combining multiple factors,
I mean, you know, thank you for walking through that.
That was really interesting. With the different timeframes, you know,
I don't see many people frame it that way, So
(18:48):
you know, can you talk us through how those different
timeframes you know, apply to a multi factor process. How
do you combine factors, especially if they have different time frames?
It is any element, and I guess, I guess I
can ask do you believe in factor timing, Like, do
you think there's a way to time factors or do
(19:08):
you think it's a better approach to just have a
relatively constant exposure to factors that you believe, you know
are are advantages for the long term.
Speaker 2 (19:21):
Sure, well, it's a two part question, So let me
take the second part of your question first, which is
we have looked for every way that we could possibly
think of time factors, and we really wish that you
could just Unfortunately there's no evidence that you can, and
that oftentimes the cost of getting it wrong is really significant.
Speaker 3 (19:44):
Right.
Speaker 2 (19:44):
One of the things I learned in engineering is you
always have to think about if a certain part or
system fails, how bad is the outcome when it fails?
And getting factor timing wrong can be a really expensive
and so generally it's better to take multiple factors that
(20:06):
are reliable but that oftentimes are not highly correlated with
each other and include them into a multi factor portfolio.
This is going to give you a little bit more
risk control and the ability to sort of ride out
or survive different periods in the market when certain factors
are in or out of favor. So I think, Chris,
(20:28):
that gets to your second question, which is, then how
do you start to combine multiple factors into a portfolio?
And I think there's a few lessons to keep in mind.
The first one that I said earlier is, you know,
more factors are not necessarily better. There's a quote that's
often attributed to Einstein that I really like that you know,
(20:51):
supposedly pretty much every quote is either attributed nowadays to
either Einstein or Mark Twain, so you never really know
if they're said, but anyway, it's a good quote which
says everything should be as simple as it can be,
but not simpler, right, So there is room for things
that are complicated in this world. But just throwing additional
(21:14):
factors into something, even though it may seem sophisticated, sometimes
actually is detrimental. And so I would say you want
to start with factors that are rigorously tested or not
duplicative with one another, and then, like you said, understand
the timeframes. And so the way that we approach it
is those long term factors that help explain returns over
(21:37):
years those are good things to actually build a strategy's
construction around, right, because you can do that in a
pretty stable way without a lot of turnover, and so
things like you mentioned earlier you were talking about that
small cap strategy, So things like valuation, profitability, size, those
(21:57):
are great to include in a long term strategy. And
the way that we do it is we will start
with market cap weights of securities, and then to the
extent that securities look good across multiple factors are bad,
we'll overweight and underweight relative to market cap weights. Now,
then this brings in some of those shorter time period factors,
(22:21):
and so with those, I think it's important to not
apply them in the construction because if a factor is
changing its signal or it's information is providing you every
couple of weeks, then it's going to cause a lot
of turnover in the portfolio. That could cause very high
trading cost or if you have taxbile investors, that could
(22:42):
cause them a huge capital gains tax bill. So there,
I think the best thing to do is to use
them as delays. So let's say that you would have
purchased or sold some security based on the long term factors,
but then you screen them for the short term factors,
and then you may decide to either delay that buy
(23:04):
or sell and then substitute in another name that isn't
having that maybe negative short run expected return. And so
I think that tends to be the way that we
think about it. And then the last thing that I
would say is you really want to understand how those
factors interrelate with one another. And so for example, value
(23:26):
and profitability, they tend to be complements. So more often
than not, when the value premium is negative, the profitability
premium is positive, or vice versa. So therefore those are
great to combine in a portfolio, whereas profitability and let's
say growth, those tend to be positively correlated. So if
you aren't careful, you could just be doubling down and
(23:49):
increasing your risk without really increasing your expected return. So
I think that's the main way that we think about
including multiple factors.
Speaker 3 (23:57):
That's so interesting. Thank you. I mean, I think it's
relative unique that you guys do like the you know,
the long term and in the short term, and I
thank you for explaining that. I mean, that's really a
really cool perspective. Who knows if Leonardo da Vinci actually
said this, But people say, Leonordo da Vinci said, simplicity
is the ultimate sophistication. That's what that's I like that.
Speaker 2 (24:18):
Yeah, that's great. So you can borrow my Roses's Day
and I'll borrow your supposedly Anaro DaVinci.
Speaker 3 (24:24):
And who knows if you said it or not. But
you know what, when you when you name drop Leonardo
da Vinci, you sounds smart. So there you go.
Speaker 2 (24:31):
Maybe it was Mark Twain, Yeah.
Speaker 3 (24:32):
Exactly, who knows it was Yogi Berra? No? Yeah, you
kind of let me do another question I had so,
you know, like I write a lot about factor investing
and and sometimes people that maybe you know, certainly aren't
as sophisticated as you and might not know much about
this stuff, they'll come back to me and say, why
(24:53):
don't you have growth as a factor? Is in growth
a factor? You know? What's the difference between something like
profitability or maybe a more broad definition you would say
quality and value? I'm sorry, and growth? Are they the
same thing? Is one better than the other? Why do
(25:13):
you always talk about quality slash profitability and our growth?
What would you say to a question like that?
Speaker 2 (25:19):
Yeah, I'm glad you bring it up because I do
think it's confusing for a lot of people, and they
often conflate these different factors, and so I see the
same thing when I talk with people, and I think
some of it just comes down to not being clear
about how these things are defined. And so for us,
(25:41):
we define value as companies that have low valuations. So
you can use various metrics. The good news is that
they all actually contain some information. But some companies have
low valuations, some have high. To us, the low valuation
or low relative price is value and the high relative
price is growth. Where when you get into quality and look,
(26:05):
I know other people have different definitions of growth, including
companies that are growing their earnings, which is also kind
of related to profitability and momentum. There's some interesting work
Robert Novi Marx who's a professor at the University of Rochester.
He's looked at momentum and profitability growth, and so there's
sort of a version of earnings momentum that's really interesting research.
(26:30):
But at least those you can tie them to specific
line items on a company's income statement or balance sheet.
I think with quality, unfortunately, I have to say it's
a bit more of a marketing term. Than a financial term.
And what I mean is I think it's designed to
appeal to people's sort of intuitive sense of, oh, well,
(26:53):
this company has quality earnings or quality balance sheet, but
there's no standard definition of that, and so I think
this causes a lot of the confusion. And when we
look at the research on quality, there's a number of
variables that managers tend to use, so return on equity, leverage, earnings, variability, others.
(27:21):
And unfortunately, when you add those factors in to a
model that already contains profitability, they don't add any additional
explanatory power. So I think for listeners, if you have
a portfolio that is already focused on valuations, profitability, and
(27:41):
then momentum considers momentum as well, you're pretty much picking
up all the effect that you're going to get from
both quality and growth.
Speaker 3 (27:51):
I couldn't agree with that more. I mean, even in
my own work, you know, I've kind of moved away
from saying quality for those exact reasons you said. People
have different ideas, and you know it does play on
people's like, of course you want high quality on low quality, right,
who wouldn't. Yeah, But I mean in my research and
I'm sure you'd agree with this, Like, profitability is by
far the biggest driver of quality, and the other things
(28:13):
really add negligible value, and so why don't we just
use profitability. It's much more easy to understand, and I
think that the you know, the research there is kind
of more clear. So I totally agree with you.
Speaker 2 (28:24):
Yeah, I agree with you, and I think Leonardo da
Vinci would agree with you as well.
Speaker 1 (28:29):
Talking about factors, you know, you mentioned long term factors
kind of you know, it's the basis for the portfolio management.
So certainly if you can kind of go into the
investment process a little bit of you know, from a
manager's standpoint of, you know, what are they? What is
the process of, you know, taking this research and implement
that into an actual portfolio or you know, our funds.
Speaker 2 (28:53):
Yeah, sure, I think I would describe our investment process
using three words. The first is stematic, the second is daily,
and the third is flexible. So let me say what
I mean by that. Every single day we take current
market prices and the most recent company fundamentals and we
(29:17):
use that to assign companies to in terms of different factors,
so value or profitability. Also, we then calculate sort of
theoretical weights for every security in every portfolio. And I'm
really gonna highlight the word theoretical there because this is
only using the long term factors, so we haven't enriched
(29:39):
that with additional information about the short term ones yet.
So it's sort of a rough work and process, if
you will. And so every day, though, we have a
description of those securities and where they sit across the
different factors, and that's very different from an indexed based approach.
(30:00):
So let me just contrast it real quick. If you're
invested in a value index, for example, or a profitability
or quality or whatever index, they're only doing those updated
calculations and bucketing of securities maybe at most every quarter. Oftentimes,
even when they say they have quarterly rebalances, they're only
(30:22):
really redefining the breakpoints between those factors on a semi
annual basis, So they're working with stale information. And so
we have a daily process to combine securities and financial
metrics categorize stocks. At that point, our portfolio managers will
(30:46):
review that updated information and where those stocks are sitting,
and we'll compare it to our current holdings, and so
then that may suggest that we may want to do
some rebalancing. So you're familiar that in X is rebalanced
maybe a couple times a year. Well, we rebalance a
little bit every single day, so it's like having two
(31:06):
hundred and fifty or so rebalance events throughout the year.
And so every day we're looking at the cash flows
coming in and out of the portfolio and thinking, how
do I use those as efficiently as possible. If there's
some security that became a lower valuation or more profitable,
and we want to increase our weight in that, how
do we use the cash flows coming in from either
(31:28):
clients or maybe from corporate actions like dividends, how do
we redeploy that cash just to the stocks that we
want to increase our weight in. And then that's when
we start to apply those shorter term criteria, those shorter
term factors, And what that will do is it will
cause us to delay from trading some of the names.
(31:50):
So we will say, all right, well, there's maybe ten
securities in the US arch cap space that we want
to buy more of, but three of them maybe have
these short term negative factors, We'll go by the other
seven instead. And from there what we do, and this
is this part becomes very unique now compared with anyone
(32:11):
else to know in the industry, is we will send
those over to our traders, and let's just say hypothetically
that we want to spend fifty million dollars, we will
give our traders, let's say three times that amount of
order candidates, So we'll give them one hundred and fifty
million in order candidates, and we will give them the
(32:31):
exact share counts and price limits and everything, so they
don't have discretion on which securities to eventually buy. But
what we do is we give them flexibility over the
timing and the quantity. So we say, buy anything off
of this list today, We'll come back and do it
tomorrow and the next day and the next day. And
so what ends up happening is our portfolio managers get
(32:54):
all the positions they want, but our traders also get
the flexibility to not have to cross spreads or push prices.
And so this helps avoid an issue that both index
funds and active managers have, which is, in some ways
most of them are demanding liquidity from the market. They're
(33:15):
going and saying I need to trade a specific stock
in a specific quantity at a specific time and when
you do that, you just don't get great prices. Whereas
if our traders can sit over on the favorable side
of the spread and let other people cross and we
can get some price improvement, that actually is a value
add in our process. So that's how the process works.
When we say it's systematic, we've built systems to do
(33:40):
this daily rebalancing in the lowest cost way that we can.
Speaker 3 (33:46):
Really interesting, it's like buying one ros a day going
up to Valentine's Day.
Speaker 2 (33:53):
Yeah, Chris, I've tried to extend this analogy where sometimes
I say, we give our traders a shopping list and
tell them to go to the grocery store, but then
we only give them a budget to buy like a
third of the shopping list, and so every day they
have to buy what's on sale. It's just at some
point you stretch the analogy so far that people are like,
who would go to the grocery store that often? But
(34:14):
with electronic trading, actually you can go to the grocery
store all the time. It's fine.
Speaker 3 (34:18):
Sure, Yeah, And I bet that working those orders, I
bet that means you know, a big difference over time.
I could totally see that, and like you said, I mean,
when you have an index, everyone knows when you're rebalancing,
and it's not you know, you can front run that stuff.
All right, let me ask you a question. This is
like as controversial as it gets, right when it comes
to factors small size. So when I first learned factor investing,
(34:42):
it was like small size as a factor. And by
a factor, I don't mean like a risk factor, I
mean like an alpha factor, like it's going to give
you higher risk adjusted returns. And then I feel like
over the last decade or two, the evidence of small
size being an actual premium has really been hit, and
some argue that it was never a premium at all.
(35:03):
It was just you're taking tail risk because you're buying
small companies and they could go bankrupt and there's more
volatility and more left tail risk there, so you should
be compensated for higher returns. It's not actually a premium.
As you know, small sizes in many you know academic
factor models. So where do you come down on this debate?
I mean, do you think small size is still a
(35:24):
factor or have you kind of reassess that over the
last couple of years.
Speaker 2 (35:29):
Yeah, So I think two things. The first is it's
always important to continuously reassess the evidence. No one should
ever just stick with what was done at some point
in the past. Being a statistics nerd myself, for those
other listeners out there that are this is called taking
(35:50):
a basin approach. It's you have your priors, you get
new data, you update you know, you add it to
the information, you update your priors. And so that's a
big part of what our research team does here is
continuously test if these things are still reliable. And then
I think the other thing to point out is in
the US we have seen a couple decades where small
(36:13):
caps have done worse than large caps, which I think
is interesting for two reasons. One is that hasn't been
the case outside the US, so across all the other
non US countries in aggregate, we've actually seen positive size premiums.
And then in the US, I think there's two things
going on. One is that people tend to use the
(36:36):
Russell two thousand as their proxy for how small caps
have done. But we talked earlier about some of the
performance drag associated with index reconstitution, and so we said
that's coming out of the index. So really that small
cap index has particularly low returns compared with some other
(36:56):
small cap indices. And then the other thing else is
you asked me earlier about multiple factors. You always have
to control for all the other So size is just
a one dimensional concept. Actually, isn't that helpful? Right? It's
kind of like if you were a medical researcher and
you said, like, did someone eat a healthy diet. Well,
(37:22):
I'm a cyclist, and so I actually eat a lot
of sugar, but I need that for exercise, and the
net effect is it's good for me in some quantity
at some time. So you always have to think about
multiple explanatory variables. And with small the issue in the
US actually has not been most small cap stocks. It's
been a very small subset of small cap stocks that
(37:45):
have very high valuations and very low or often negative profitability.
And so some people like to sort of casually say,
you know, there's a lot of junk and small caps,
and so I think in small cap investing you have
to be very careful that you are controlling for those
other factors as well. Once you do that, yeah, there
(38:07):
is a premium for small cap stocks, and we tend
to see that some of the other premiums are actually
a little bit stronger in small cap than in large.
Speaker 3 (38:15):
Yeah, I found that too, of the other factors, and
you know, just simply doing a profitability screen on the
Russell two thousand goes a long way, you know, like.
Speaker 2 (38:24):
Absolutely, yeah, absolutely, I mean it's sort of we like
to geek out about factors and be quantitative, but just
step back common sense. Wise investing is always about what
am I paying for something versus what am I expecting
to get right, And if you pay a high valuation
(38:45):
for something that has little to no profits, that's not
generally regarded as a high expected return investment. And so
I think it's always important in any area of the
market to use that framework, but especially in small caps
because there are just a lot of those unprofitable small
cap companies in the US.
Speaker 3 (39:05):
All right, let me ask you about kind of the
topic of the day, which is AI machine learning. You know,
you you mentioned deep seek. That was the you know,
big topic of the markets yesterday. Have you found you know,
applications of it could be just mL models or even LM,
you know, AI models in factor investing or if you know,
(39:28):
if not, like, do you think that's a thing we're
going to you know, really focus on in the future
trying to have machine learning and such help us with
the factors. What's your thoughts there.
Speaker 2 (39:40):
Yeah, it's something we've looked into a lot. And the
interesting thing you mentioned deep Seek again is they actually
started out as a quantitative hedge fund, and that hedge
fund ran into some troubles and so they've sort of
pivoted over to a generative AI model. And I think
it just highlights that it's been very difficult to apply
(40:02):
AI to try to whether it's time the market or
find new factors. In some ways, you start to worry
about what researchers would call data mining, which is you know,
we talked about the four hundred factors and the factor zoo. Well,
you could probably enumerate hundreds of thousands or millions of
(40:23):
combinations of different variables if you were to combine things
off of the income statement and the balance sheet, so
you would make all sorts of esoteric variables and then
you would run them all through a factor model and
by chance some would work right. So this is called
pa hacking in the statistics community. It's one of those
things where you know, even if there's only a five
(40:45):
percent chance that something happens by random. If you run
one hundred tests, well, then five of them are going
to turn up positive, even though they're just by chance.
So i'd mentioned Robert nobe Marx. He actually wrote a
paper on this topic recently. He not only only created
a bunch of AI models to find new factors, he
also then employed AI models to write the academic papers
(41:08):
for him. And it was done as sort of a
way to demonstrate to the industry what can go wrong
if you use this approach. So it is something we've
looked at a lot so far. The ways that we
have experimented with using it is more to gather some
of that unstructured data that may be like in company
(41:29):
filings so or oftentimes when companies have corporate actions, they're
putting out a bunch of text based information. And if
you're using different factors, so for example, if you're using
a profitability factor a value factor, it's relying on the
numbers in the income statement to be interpreted properly. And
so let's say that you know the auditor qualifies their
(41:53):
opinion about the income statements, Well, do you want to
trust the metrics from there? Probably not. So if you
had some sort of a model that could scan through
and highlight to you areas where there may be concerns
or maybe changes. So let's say a company spins off
a division. Well, now anything on the balance sheet about
you know, assets or book value of equity has changed
(42:15):
because they've spun some of that off. So finding ways
to flag some of those changes in the data so
you can go update your metrics so you're not using
stale metrics has been an area where we've experimented with
using it.
Speaker 1 (42:29):
Oh, this is great. We just have one more question
before we let you go. You know, like to ask
this a lot to a lot of our guests. But whatever,
you're some of your favorite financial or investing books.
Speaker 2 (42:40):
Yeah, so I read financial literature all day long at
my job, and so when I get a chance to
read books, I tend to like books that come from
other fields. And I really enjoy history. I think we
have a lot to learn from history, and specifically I
like biographies, and so one of my favorite authors is
(43:02):
Walter Isaacson, and he's written biographies on people like Steve Jobs,
Leonardo da Vinci, which you were talking about earlier, Elon Musk,
Benjamin Franklin, and others. And one of the things I
really like reading about in these biographies is each of
these individuals accomplished a lot. They were all great problem solvers,
(43:27):
and each of them went through a lot of challenges.
They went through big ups and downs along the way.
And I think intellectually we all know that success, whether
that's in life or in investing, it's not a line
that goes straight up into the right right. But there's
a difference between knowing that and then actually being able
(43:48):
to have the discipline to sort of live through the
downs and capture the ups. And so for me, reading
biographies really brings to life the lessons from history, so
I can learn from them and not have to repeat
them myself. It's great, Joel.
Speaker 1 (44:04):
I enjoyed this. Thanks again for joining us today.
Speaker 2 (44:06):
Absolutely, Thank you for having me and Chris.
Speaker 1 (44:08):
Thank you for being my co host again today.
Speaker 3 (44:11):
Thank you, and thank you Joel.
Speaker 1 (44:12):
Until our next episode, this is David Cohne with the
Inside Out.
Host
Gina Martin Adams
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!