March 27, 2021 • 29 mins
This week, Jason Tuvey, Capital Economics senior emerging markets economist, joined to talk about the sharp selloff in Turkish assets after President Erdogan ousted the country's central bank governor and what comes next for the country's economy and financial markets. Bloomberg Opinion columnist Noah Smith explained why local price changes aren't inflation. Harvard Business School professor Willy Shih discussed Intel's ambitious bid to regain its manufacturing lead. Then Sridhar Ramaswamy, a former senior vice president at Google, who ran the company's ad business for fifteen years, reacted to the House tech hearing and discussed whether these companies' business models will ever be put under threat.
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 (00:08):
Welcome to Would You Miss This Week? I'm Joe Wisenthal.
This podcast has some of our favorite interviews from the
Daily Market Clothes show that I co anchor along with
Romaine Bostick and Caroline Hyde. What do you Miss? It's
the perfect way to kick off your weekend. This week
started off feeling like deja vu in Turkey, Turkish stocks, bonds,
and Dealira tumbling in one of the sharpest selloffs in years.
(00:29):
That coming on the hills of President Ridwan, We're moving
the country's central bank governor after only four months on
the job. The new governor is a fan of lower
rates like Alan and marks the country's fourth central banker
in less than two years. So we spoke about what
comes next with Jason Tuvy, a senior Emerging markets economist
at Capital Economics, and we started by asking Jason, what
long term impact this ongoing uncertainty we'll have on Turkey's
(00:53):
economy and financial markets. Yeah, I mean the moves today
we're critine credible, to be honest, and I mean under
the previous governor that's abound. There were signs that maybe
Turkey was turned in a corner. Maybe this it looked
like this policy shifted we've seen since November was the
(01:13):
real deal, but clearly give up his unorthodox theories for
too long and a Rugby fired Mr agbo U. So yeah,
we now want to another central bank governor, it seems,
someone who is more aligned with what President Dan wants.
(01:34):
I mean, the result is likely to be some effort
at some point in the near future to bring mates
down again. But that's something You're going to make Turkey's
inflation problem much worse, and that probably means continued weakness
in the leer we're going forward. I mean, you saw
serious knock on effects sort of Globally speaking, we saw
(01:56):
Spanish banks take a massive hit. Those most exposed to
Turkish loans in new are up in particular. Still, that
fine line can be drawn between the two. We'll talk
to us about the Turkish banking sector and how much
that is a key concern for you. Yeah, I think
the Turkish banking sector is the key area of a
vulnerability in Turkey's economy. Now, there are some concerns about
(02:17):
the effects of the lear in terms of mismatches on
banks band sheets, they're not those concerns aren't actually as
big as they were, say back in last summer. But
the concern that's been there throughout the past few years
has been the fact that banks have very large external
debts and these have to be repaid world over quite regularly.
(02:41):
These amount about twelve and a half percent of GDP.
And to put that into context, banks um the liquids
assets at the central Bank which they used to repay
these these debts during periods of stress. These are these
assets are very very low um. And what's more, the
Central Bank has heavily depleted its foreign exchange itself, so
(03:04):
it can't step in either if if banks do got
into trouble UM. So I think we will seeing seeing
the lever for today Cdspringer jumped on upon spreads wide
and sharply. I think it is the banking sects where
we really need to keep a close eye on the
coming over the coming days and weeks. So what in
terms of the future of the Central Bank going forward
(03:26):
and what is what do we likely to see in
terms of our policy, both on raids policy but also
as you mentioned, the sort of like domestic policy to
ensure you know, holding vira. Yeah, it's a good question.
I think in terms of rates it seems ely one's
desire clearly is to bring lights down. He has been
an advocate opponent of PI interest, right, So so that's
(03:49):
the desire getting too. Given the very poor banks, appointment's position,
I think is going to be very difficult in terms
of trying to get people to tolbably live. I think
inevitably there will have to be some sort of capital
controls put in place. Maybe not a very immediate future,
but I think that's the road that Turkey is heading
(04:10):
to heading down in the long run. Trust a little
bit more about the global ramification of this. I mean
I talked about how European banks got drag lower. But
as we do go into day two to digest this,
who else gets hurt? Yes? A good question. Actually, I
think I think the risks are really contained two Turkey.
This is such a idiosnocratic thing going on that I
(04:35):
think we seem period for say, following the country crisis
in where there were some hits hit to say other
M currencies due to a period of risk of sentiment.
But I think we've seen today actually the moves in
the M currencies, there's barely been any. I mean the
rand and the liver were down one episode. Demand and
(04:57):
the Mexican past over down about one percent against a
lot of earlier in the day, but they quickly snapped back,
and I think investors are taking this is this is
purely a Turkey story. What do you watch for though
in terms of identifying vulnerability? So right now it's purely
a Turkey story, but from a fundamental standpoint, and again,
(05:18):
we do seem to be in this pretty up swinging
markets and despite the raids, UM seems to be doing fine.
But what are the radar signs you look for of
when E m's perhaps get into trouble. Yeah, I mean
I said Turkey is very particular here. I think the
external position in particular if you look at that current
(05:40):
account depity is really large, got large external debts, and
the effects reserves Turkey on on those measures that sort
of out on the league of its own UM. But
we would be looking at similar things for for other
companies to try and assess vumabilities. And I think that
you mentioned any specific link to Turkey, we would be
(06:01):
looking forward, such as that you said the Spanish banking sector.
But beyond that, I think, um, well, I think most
of the vulnerabilities in EMS at the moment are quite
domestic driven in Brazil and violus developments and what's happening
with the public finances there are simibly in South Africa
with public finances there, and I think it's very difficult
(06:24):
at the moment to draw specifically and themes beyond what
we see in inside the US Treasury mark and any
mamnifications there. Now this week, there's been plenty of discussion
about the eye word as the debate over inflation rages on.
Bloomberg Opinion calumnist Noah Smith came on to talk about
why your local price changes aren't inflation and why some
(06:47):
people in the Bay Area may think inflation is high
even when it's not. We started with the empirical side
of the debate. Nest no if we should be worried
about what we're seeing the producer price levels and whether
they'll feed through to consumers. Well, if feed through, probably yes.
I mean, you know companies try to pass on increased
cost to their consumers. Uh, we'll lead to something dramatic.
(07:09):
I don't think we know yet. I think if that
would be like a multi month, multi quarter, you know,
multi point rise in inflation. At that point we start
to freak out and say, okay, what the heck is
going on here? But we're just not there yet, not
there yet. You've got a great piece out today now
talking about Lucas Islands model. I mean, well, in what
(07:33):
perspective we want, who is measuring inflation? How are they
measuring it? And we will measuring it in the same ways. Right,
So that was about Lucas Islands model was this old
theory about how everybody tried to infer inflation from how
much they saw their local prices go up. And I
think that a lot of people still do this. You
you know, I have some friends living in the Bay
Area who say, oh my god, houses this expensive, rented
(07:54):
this expensive. Uh, you know, restaurants are this expensive. We
must have a lot of an inflation. But in fact
that's just local. That's because a bunch of rich people
moved to the Bay Area and if you go to
you know, Cleveland or wherever, that's just not happening. And
so you have to look at the national level. So
the other thing that a lot of people in the
Bay Area, and I have to say it seems like
(08:16):
this year people in the Bay Area all started getting
opinions on inflation. Another thing that they say a lot
is that we should talk about like asset inflation, that
for somehow stocks and the cost of art or the
cost of an n f T is going up a lot,
and therefore that has something to do with inflation. Why
is this sort of logically not something that we should
think about as inflation the way the FED thinks about it. Right, So, basically,
(08:40):
inflation is meant to be kind of, you know, the
change in the value of the currency. So the currency,
you know, the actually US dollar is worth less relative
to all the useful things you might want to buy,
such as you know, bread and uh, you know, back
massages or whatever you want to buy. I guess not
in the pandemic. But but so, so the thing is
(09:00):
that a stock isn't actually something useful that you want
to buy, because no one actually needs a share of apple.
They just, you know, apple shares are just sort of
acclaim on the useful stuff that's up. So because apple
shares don't have intrinsic value, we don't want to include
them in inflation. That doesn't mean that asset price changes
aren't important. It just means that that's not what we
think of when we think about inflation. So what economists
(09:26):
J Pow out there today not worrying about inflation, obviously
desperate to see it. Go about that two level for him.
We're economists seeing inflation at and what are they using
to measure? Is the right thing from your perspective? Well,
I mean, um, you know the all the measures of inflation,
you know, the c P I, the GDP deflator, the
(09:48):
producer price indusicries um PC, and even the Billion Price
Project by m I T. These all are very highly correlated.
Over the long term, there can be some divergence, so
we argue about which one we should using. We're measuring
how much stuff has changed since like, but if we're
talking about how much stuff changed this month. These these
are all pretty much correlated, and they're all sort of measuring,
(10:10):
you know, the same stuff and um, and of course
you know the FED is going to be looking at
them all, but they're all saying the same thing. So no,
it's something else that you wrote about today on your newsletter.
It's like the return of macro wars, this idea of
different ideological camps having a lot of spirited people like
Larry Summer's warning of overheating the m MT types much
less concerned about it. A lot of it comes down
(10:32):
to this question of do we even have good models
for reliably anticipating inflation? And it feels like the FED
is sort of like, we actually really don't know when
we're gonna get inflation, so we'll just wait till we
see it. How well, in your view, have the traditional
approaches held up and do we just have to like
do we just have to jump the old ideas of
where we think we know what causes inflation? I would
(10:54):
say that, um, macro models of the kind of academics
use and the kind that they try to get the
to use have been completely useless and predicting inflation, um,
just just completely useless, and that at this point everyone
is just relying on kind of heuristics and rules and
kind of Kinesian ideas that people learned an icon one
oh one. Uh you know either that, and at the
(11:16):
level of the public discussion it's even less. It's just
people you know, shouting like mm T memes and stuff.
But then, um, but I would say that in terms
of people at the FED, they're using these simple heuristics
and rules because academic theories there's no help. And what
is when inflation you actually see it. So yes, you're
gonna have to be in the pc data. But is
(11:38):
it going to come down to wage inflation that you
eventually have to see if we're going to see anything sustained?
And at what point does that happen Apart from in
the Bay Area, maybe maybe not so. An interesting thing
about wages is that, uh, when you have moderate inflation,
you know, workers are able to bargain for raises, like
cost of living raises. But then when inflation gets really
high and variable, you know, where it gets up to
(11:59):
like eight or ten percent, but then also bounces around,
workers have a much harder time, you know, anticipating how
much inflation will be and then bargaining to keep up
with that. And so therefore wages you know, often, you know,
at least in the seventies, you know, wages fell behind
during inflation, which doesn't happen in your textbooks but kind
of happens in real life. And so, um, you know,
(12:22):
they're not counting the CPI. Maybe there's some arguments about
that they should be, but the fact is that it's
not wages that's going to show us. It's just you know,
consumer prices, producer prices and stuff like that. So from
your perspective, just to tie it up in a bow,
do you think we've got inflation or not? It's too
early to tell, you know, give it, give it a
few months. Then these things bounce around a lot. It
(12:43):
could be gone next month. This week we also got
a major announcement from Intel, the company's new CEO, Pat Gelsinger,
unveiling an ambitious bid to regain its manufacturing lead by
spending billions of dollars on new factories and creating a
foundry business that will make chips for other companies. Now,
this was not the plan Wall Street had been expecting.
(13:05):
Many thought Intel would go with an ultra light asset
model designing chips and outsourcing, being the capital intensive manufacturing.
We spoke about it with Willy She, Professor Management Practice
at Harvard Business School and an expert in manufacturing and
product development. We started by asking Professor She, just how
difficult this plan is actually going to be for Intel
to execute? Well, uh, twenty billion, that's a start, okay,
(13:32):
I think, uh, you know, it took years for them
to fall behind, Okay, And this is a tough problem.
This is tough technology. Okay. So it's a start, but
clearly Pat Gelsinger has put down a marker and saying
we're climbing back into the ring here and we're gonna
go after leadership again. So you know, I think that's encouraging.
(13:54):
What about where they're doing it in terms of is
it rightly sil SMC and there's an asia full? Is
it right to be focusing so much on what's happening
in Arizona for example? Over in Europe as well well?
I think these shortages that we're seeing, particularly in automotive
but increasingly in other sectors as well, has called attention
(14:16):
to kind of the fragility of supply chains, of vulnerability
the uh kind of excessive dependence on a couple of
suppliers foundries in Taiwan, okay uh. Now, so you see
a lot of that tension in Washington about how we
need to have more of that supply chain domestically. Now,
it's going to take a while to rebuild that. Intel
(14:39):
is proposing to go into the foundry business, I mean,
and they're making a big bet on that. It's going
to take them a while to develop that model. Now.
The good thing is they're starting with two new fabs
that are going to be dedicated to the foundary business.
And what Pat Gelsinger has said is they're going to
do industry to standard tools. They're going to use industry
(15:02):
standard uh I P blocks. Those are the chunks of
circus that you use uh to develop those chips. Right,
So it's going to be more aligned with the model
that other customers who go to the time when these
foundries like t SMC are used to doing. It's not
a model that Intel is used to doing though, So
you know, twenty billion dollars, it's a lot, but even
(15:24):
you know, back in January, t SMC announced it was
going to spend another twenty billion on KPEX. So they're
kind of table sticks for these industries, at these players
to spend a ton of money. When I asked you
is this gonna work, You're like, well, it's a start.
What's it gonna take beyond this initial commitment of an
investment for them to reverse what, as you said, was
(15:44):
like a multi year process of falling behind. Well, you
know what they have to do is they have to
really push the frontiers. Now, what Pat Galsinger has said
is they're going to use they're going to commit to
deep u vy. They've been a little behind on that
right that that, by the way, is one of the
reasons for the twenty billion dollar price tag because when
you look at one of these sml dpub UH machines,
(16:07):
you know, each machine is pushing a hundred fifty million
dollars and in a fab the size that Intel is
going to propose the build, they might need twenty of them. Okay,
So that's like having a fleet to these, you know,
a hundred million plus machines in there. Okay. Now, what
they need to do is they need to get the
(16:28):
processes up and they need to get volume through there, right,
because it's through volume that's how you perfect your processes,
that's how you do your learning, that's how you drive
your costs down. Okay, And t SMC is very good
at that. Of course, they've spent thirty five years developing
that skill. Intel knows how to do that with their microprocessors,
(16:49):
but they've had some troubles with advanced processes, right, so
they'll they'll they'll have to work at it, and it's
going to be multiple years. It's not something UH that
will be able to look two quarters, three quarters, or
even two years out and say, Okay, they're going to
solve this problem. It's going to be a steady progress
(17:10):
and they're just gonna have to take risks and they're
gonna have to make investments. They're gonna have to get
customers to buy that output h so that they can
keep driving volume and learning and almost that exact thing
that it's not going to be solved in a couple
of quarters, that they're going to have to keep on reinvesting.
Can we said exactly the same for the shortage in
chips that we're seeing at the moment, and just take
(17:32):
a listen to what part Els think I had to
say about that. COVID caused everybody to step back a
bit from capacity and build out and the supply chain
the industry, and it induced a radical increase in demand.
So you have supply chains, uh scaling back a bit
and demands scaling up radically. Wow, and now we're in
a position that there's a meaningful shortage and you know
(17:54):
it's going to be a couple of years until that's
fully resolved. Will you have to go on on that
global shortage that we're seeing at the moment and how
that gets resolved? Yeah, Well, there are a lot of
interesting fascests to the shortage. Okay, and we're feeling in
an automotive. But I will remind everybody that you know,
(18:14):
at the beginning of UH, the outlook for the global
automobile market, including China in particular, was slightly down from
UH from two thousand nineteen. And that was before COVID hit, right,
So then COVID hit and sales fall off a cliff UH.
And so if you're an automaker, then you're saying, how
I'm gonna survive, How I'm gonna I'm gonna shepherd cash
(18:36):
and so on, So they pulled back orders. Now, with
the pandemic, did IS created huge demand for things like
note but computers, flat panel displays, consumer electronics, new gaming consoles.
We had new generations that gaming consoles come out. Okay,
and you know, to the extent of the capacity is fungible.
They soaked up a lot of the capacity. So meanwhile,
(18:57):
the auto guys and this started to become visible late
last fall November, it was already starting to become visible.
The lead times for getting your orders into the boundaries
started extending out Now, the US worse by sanctioning Huawei
and sanctioning s M I C so Huawei put in
(19:17):
a lot of orders for chips last summer in Taiwan
and that absorbed a lot of capacity, right, So it
was it was a combination of factors. Now to get
through this uh shortfall, Now, it's gonna take. It's gonna
take a good six months a year before we really
start seeing a lot of relief on that. And we
(19:43):
ended with yet another big tech hearing on Capitol HILP three,
prominent Silicon Valley c e O s Soon Arch Google,
Mark Zuckerberg of Facebook, and Jack Dorsey of Twitter facing
questions from House lawmakers over disinformation on their platform, particularly
in reference to the January insurrection at the Capitol. What's
which was organized online? We've gotten used to these hearings
(20:03):
and they often are not very enlightening or market moving.
But we caught up with Srida Ramaswami, a former senior
VP at Google where he ran their ad business for
fifteen years, to ask whether they will ever be a
meaningful policy turn to open these companies business models and
the threat. Well, I think we need to start with
just having more information part of what is disappointing about
(20:26):
today and honestly about the last three months is we
don't know that much more than what we knew on
January six. Think about it, Joe, if January six had
been an airline accident, the NDSB would have been all
over the place. We would have actual data about what
actually happened. Um while with the tech platforms, we don't
really know how good they were at, say to moving
(20:47):
even illegal content. We're not even talking about things like misinformation.
So I just think we need to be much better educated.
And and you know, our government needs to set up
something like the NDSP to oversee this stuff. If he
can receive airlines, why not orse social media? Only then
will we have meaningful change. Otherwise, you know, we sort
of have this political theater that doesn't really go anywhere importantly.
(21:10):
You say that, of course, as you were overseeing Google's
advertising and commerce products that ended up being a hundred
billion dollars revenue run rate when you were there, Street
and I'm interested you've gone on too, of course, be
a CEO of an add free personal consumer facing search service. Instead.
Your view therefore on on really the regulatory impact of
(21:32):
what you hope to come out of this. Do you
want to see more regulations, just saying, as an oversight boarder,
is there something more deeply that needs to be changed
within section? I'm just saying that oversight more information will
just let us be a lot more educated about what
kind of changes to things like section to thirty are
(21:53):
reasonable to make. I've argued, for example, that we need
to separate things. Downs need to separate out what I
call organic content, content that say, you are Joe share
with your followers on Twitter, versus stuff that's algorithmically amplified
by the platforms. The latter, in my opinion, should be
held to a much higher standard. And the one thing
(22:14):
that they never talked about is all platforms have the
option of delaying content if they're unsure about its nature. Um,
so these are things that they could be doing. Of course,
what I'm doing is completely you know, different from any
of the business models that are out there. Uh nivas
An adds free private search engine because we think having
you as both to use it on customers will just
(22:35):
lead to a better experience over time. So you know,
on that you talk about the distinction that you established
between more organic sharing and sort of more algorithmic things
designed to go viral. I mean people post a lot
of nonsense on different social media platforms, and they often
don't do it as part of some coordinated campaign. They
(22:56):
just tell their friends some misinformation that they heard about
vaccine means or something that they heard about voting machines
or something like That doesn't mean that they like got
it from on top or someone was telling them to
do that. Um. Should the should the platforms think about
throttling that which is just me talking to you excepted um,
(23:16):
you know, completely made up? It's one of the tools
that they have to consider. If it is you know,
our proverbial crazy drunken uncle talking to do office friends,
you know, maybe nothing happens to that. If on the
other hand, if it is a powerful politician with a
megaphone of fifty million people, their words are just completely different.
I think holding that to a higher standard is fine.
(23:39):
Guidelines from Congress, guidelines from regulators are going to help.
But I think we do need to understand that those
two just are not the same. What you and I
might say in the confines of our home to a
couple of friends is completely different from what somebody says
to a hundred million people. Should know how far can
AI go at the moment as it stands, because there's
(23:59):
been in full lot. You hair a lot from Facebook
as the amount of people they have working on this
human eyeballs, but also how they want artificial intelligence to
really be able to pick out the most damaging misinformation,
take it offline as quickly as possible. How good is
the AI getting how quickly can it be enforced? You
think I've dealt with some of the hardest problem at
(24:20):
scale when I was running ADS. AI is not a
magic answer to all solutions, UM, and you run into
the kind of constraints that Joe is talking about, which is,
if you're super accurate at identifying what you think is illegal,
you're going to sweep up a bunch of pretty regular
content along with it. This is why I'm saying we
need more data. We need to know things like if
(24:42):
illegal content is shared on such and such a platform,
they can act on it within two minutes. Are they're
going to delay it for twenty minutes while actual people
look at this. UM At the end of the day,
we need to care about more than effort. We need
to have a shared understanding of the actual thing things
that happen, of actual results. So in my mind. Just
(25:03):
saying we have ten thousand people or twenty thousand people
doesn't let you or me know are they really good
at solving this problem? Yeah, it's part of the answer,
but AI has more constraints than people really want to
talk about. As an investor, thinking about the competitive landscape
for tech, how do you think about the difficulty in
funding companies that theoretically compete with these uh gigantic firms.
(25:27):
How much does that enter into your competition and do
you think that there is a competitive marketplace for the
sort of consumer or other Internet things that that interests you.
It's pretty hard. I think you know, investors understand that
going head to head against very large companies is very problematic.
(25:48):
On the other hand, enterprise investing, that is a there
is a lot of the going on fintech investing, there's
a lot of the going on. In fact, I would
say that NIVA is one of the few pure consumer
more companies that have been funded by vcs in the
recent past um and that's because you know, there is fear.
But going head to head against brilliant dollar companies is
(26:10):
not a smart way to make money. So with your
CEO hat on, your co found a hat on of
NEVA and and being in that competitive space to a
certain degree, Siata, what is the data that you say
you want to see? You say you need more data
to really understand what sort of a job, how good
a job that is they're doing. What data could you
get from these companies? Well, when I when I speak
(26:31):
about social media and data, I'm speaking about it more
as what is the public need or what is the
Congress need in order to reach better decisions? Um? And
to me, as I said, it's it's things like how
quickly subjectionable content brought down, how much of that content
is subject to review before cluse. What's the impact of,
as I said, algorithmic amplification on how much objectionable content
(26:55):
gets shared? Um? Within the context of NEIVA, we solve
a different problem with the different set of principles. UM.
So you know, this kind of data is less of
in the context of my being the CEO of Shouldn't.
But on that note, you say, would they have gladly
give that data? Do you think how would they measure
(27:16):
price to be able to measure those sorts of bits
of data that you're saying that Facebook could give to
the body politics, every team that works on objectionable content
has to grade itself internally in terms of how good
a job they're doing. The metrics that I am talking about,
how quickly subjectionable content identify, how much of it is
(27:40):
sent for human review, how are the human raiders? These
are all things. Yes, the methodology will be different from
company to company, but they have the data, um. But
there needs to be an amount of pressure to have
this data come out in the open so we can
actually make better decisions about what's like, what's the long
(28:00):
term of Section to thirty, what provision to be changed?
How would a job really are we doing? And is
that something that you think, um, the industry should be
more forthcoming with unilaterally on that or is that something
that a regulator or perhaps a legal act could could Enfortunately,
we could actually as you say, good to answer to
see these questions given the past forty years we've been
(28:22):
talking about this since the election, right, Um, I would
actually say that it's going to take some kind of
an act of regulation, perhaps the FTC stepping up to
do the job. I think it's going to require a
certain amount of pressure before more of this information is
is available freely. And that's it for what you missed
(28:46):
this week. Do you like the show and make sure
to subscribe and rate us. An Apple podcast or wherever
you listen to podcast, you can catch our show every
weekday from three thirty to five pm on Bloomberg TV
and from four to five pm on Twitter. Thanks are
listening and have a great week. M H.
Welcome to Would You Miss This Week? I'm Joe Wisenthal.
This podcast has some of our favorite interviews from the
Daily Market Clothes show that I co anchor along with
Romaine Bostick and Caroline Hyde. What do you Miss? It's
the perfect way to kick off your weekend. This week
started off feeling like deja vu in Turkey, Turkish stocks, bonds,
and Dealira tumbling in one of the sharpest selloffs in years.
(00:29):
That coming on the hills of President Ridwan, We're moving
the country's central bank governor after only four months on
the job. The new governor is a fan of lower
rates like Alan and marks the country's fourth central banker
in less than two years. So we spoke about what
comes next with Jason Tuvy, a senior Emerging markets economist
at Capital Economics, and we started by asking Jason, what
long term impact this ongoing uncertainty we'll have on Turkey's
(00:53):
economy and financial markets. Yeah, I mean the moves today
we're critine credible, to be honest, and I mean under
the previous governor that's abound. There were signs that maybe
Turkey was turned in a corner. Maybe this it looked
like this policy shifted we've seen since November was the
(01:13):
real deal, but clearly give up his unorthodox theories for
too long and a Rugby fired Mr agbo U. So yeah,
we now want to another central bank governor, it seems,
someone who is more aligned with what President Dan wants.
(01:34):
I mean, the result is likely to be some effort
at some point in the near future to bring mates
down again. But that's something You're going to make Turkey's
inflation problem much worse, and that probably means continued weakness
in the leer we're going forward. I mean, you saw
serious knock on effects sort of Globally speaking, we saw
(01:56):
Spanish banks take a massive hit. Those most exposed to
Turkish loans in new are up in particular. Still, that
fine line can be drawn between the two. We'll talk
to us about the Turkish banking sector and how much
that is a key concern for you. Yeah, I think
the Turkish banking sector is the key area of a
vulnerability in Turkey's economy. Now, there are some concerns about
(02:17):
the effects of the lear in terms of mismatches on
banks band sheets, they're not those concerns aren't actually as
big as they were, say back in last summer. But
the concern that's been there throughout the past few years
has been the fact that banks have very large external
debts and these have to be repaid world over quite regularly.
(02:41):
These amount about twelve and a half percent of GDP.
And to put that into context, banks um the liquids
assets at the central Bank which they used to repay
these these debts during periods of stress. These are these
assets are very very low um. And what's more, the
Central Bank has heavily depleted its foreign exchange itself, so
(03:04):
it can't step in either if if banks do got
into trouble UM. So I think we will seeing seeing
the lever for today Cdspringer jumped on upon spreads wide
and sharply. I think it is the banking sects where
we really need to keep a close eye on the
coming over the coming days and weeks. So what in
terms of the future of the Central Bank going forward
(03:26):
and what is what do we likely to see in
terms of our policy, both on raids policy but also
as you mentioned, the sort of like domestic policy to
ensure you know, holding vira. Yeah, it's a good question.
I think in terms of rates it seems ely one's
desire clearly is to bring lights down. He has been
an advocate opponent of PI interest, right, So so that's
(03:49):
the desire getting too. Given the very poor banks, appointment's position,
I think is going to be very difficult in terms
of trying to get people to tolbably live. I think
inevitably there will have to be some sort of capital
controls put in place. Maybe not a very immediate future,
but I think that's the road that Turkey is heading
(04:10):
to heading down in the long run. Trust a little
bit more about the global ramification of this. I mean
I talked about how European banks got drag lower. But
as we do go into day two to digest this,
who else gets hurt? Yes? A good question. Actually, I
think I think the risks are really contained two Turkey.
This is such a idiosnocratic thing going on that I
(04:35):
think we seem period for say, following the country crisis
in where there were some hits hit to say other
M currencies due to a period of risk of sentiment.
But I think we've seen today actually the moves in
the M currencies, there's barely been any. I mean the
rand and the liver were down one episode. Demand and
(04:57):
the Mexican past over down about one percent against a
lot of earlier in the day, but they quickly snapped back,
and I think investors are taking this is this is
purely a Turkey story. What do you watch for though
in terms of identifying vulnerability? So right now it's purely
a Turkey story, but from a fundamental standpoint, and again,
(05:18):
we do seem to be in this pretty up swinging
markets and despite the raids, UM seems to be doing fine.
But what are the radar signs you look for of
when E m's perhaps get into trouble. Yeah, I mean
I said Turkey is very particular here. I think the
external position in particular if you look at that current
(05:40):
account depity is really large, got large external debts, and
the effects reserves Turkey on on those measures that sort
of out on the league of its own UM. But
we would be looking at similar things for for other
companies to try and assess vumabilities. And I think that
you mentioned any specific link to Turkey, we would be
(06:01):
looking forward, such as that you said the Spanish banking sector.
But beyond that, I think, um, well, I think most
of the vulnerabilities in EMS at the moment are quite
domestic driven in Brazil and violus developments and what's happening
with the public finances there are simibly in South Africa
with public finances there, and I think it's very difficult
(06:24):
at the moment to draw specifically and themes beyond what
we see in inside the US Treasury mark and any
mamnifications there. Now this week, there's been plenty of discussion
about the eye word as the debate over inflation rages on.
Bloomberg Opinion calumnist Noah Smith came on to talk about
why your local price changes aren't inflation and why some
(06:47):
people in the Bay Area may think inflation is high
even when it's not. We started with the empirical side
of the debate. Nest no if we should be worried
about what we're seeing the producer price levels and whether
they'll feed through to consumers. Well, if feed through, probably yes.
I mean, you know companies try to pass on increased
cost to their consumers. Uh, we'll lead to something dramatic.
(07:09):
I don't think we know yet. I think if that
would be like a multi month, multi quarter, you know,
multi point rise in inflation. At that point we start
to freak out and say, okay, what the heck is
going on here? But we're just not there yet, not
there yet. You've got a great piece out today now
talking about Lucas Islands model. I mean, well, in what
(07:33):
perspective we want, who is measuring inflation? How are they
measuring it? And we will measuring it in the same ways. Right,
So that was about Lucas Islands model was this old
theory about how everybody tried to infer inflation from how
much they saw their local prices go up. And I
think that a lot of people still do this. You
you know, I have some friends living in the Bay
Area who say, oh my god, houses this expensive, rented
(07:54):
this expensive. Uh, you know, restaurants are this expensive. We
must have a lot of an inflation. But in fact
that's just local. That's because a bunch of rich people
moved to the Bay Area and if you go to
you know, Cleveland or wherever, that's just not happening. And
so you have to look at the national level. So
the other thing that a lot of people in the
Bay Area, and I have to say it seems like
(08:16):
this year people in the Bay Area all started getting
opinions on inflation. Another thing that they say a lot
is that we should talk about like asset inflation, that
for somehow stocks and the cost of art or the
cost of an n f T is going up a lot,
and therefore that has something to do with inflation. Why
is this sort of logically not something that we should
think about as inflation the way the FED thinks about it. Right, So, basically,
(08:40):
inflation is meant to be kind of, you know, the
change in the value of the currency. So the currency,
you know, the actually US dollar is worth less relative
to all the useful things you might want to buy,
such as you know, bread and uh, you know, back
massages or whatever you want to buy. I guess not
in the pandemic. But but so, so the thing is
(09:00):
that a stock isn't actually something useful that you want
to buy, because no one actually needs a share of apple.
They just, you know, apple shares are just sort of
acclaim on the useful stuff that's up. So because apple
shares don't have intrinsic value, we don't want to include
them in inflation. That doesn't mean that asset price changes
aren't important. It just means that that's not what we
think of when we think about inflation. So what economists
(09:26):
J Pow out there today not worrying about inflation, obviously
desperate to see it. Go about that two level for him.
We're economists seeing inflation at and what are they using
to measure? Is the right thing from your perspective? Well,
I mean, um, you know the all the measures of inflation,
you know, the c P I, the GDP deflator, the
(09:48):
producer price indusicries um PC, and even the Billion Price
Project by m I T. These all are very highly correlated.
Over the long term, there can be some divergence, so
we argue about which one we should using. We're measuring
how much stuff has changed since like, but if we're
talking about how much stuff changed this month. These these
are all pretty much correlated, and they're all sort of measuring,
(10:10):
you know, the same stuff and um, and of course
you know the FED is going to be looking at
them all, but they're all saying the same thing. So no,
it's something else that you wrote about today on your newsletter.
It's like the return of macro wars, this idea of
different ideological camps having a lot of spirited people like
Larry Summer's warning of overheating the m MT types much
less concerned about it. A lot of it comes down
(10:32):
to this question of do we even have good models
for reliably anticipating inflation? And it feels like the FED
is sort of like, we actually really don't know when
we're gonna get inflation, so we'll just wait till we
see it. How well, in your view, have the traditional
approaches held up and do we just have to like
do we just have to jump the old ideas of
where we think we know what causes inflation? I would
(10:54):
say that, um, macro models of the kind of academics
use and the kind that they try to get the
to use have been completely useless and predicting inflation, um,
just just completely useless, and that at this point everyone
is just relying on kind of heuristics and rules and
kind of Kinesian ideas that people learned an icon one
oh one. Uh you know either that, and at the
(11:16):
level of the public discussion it's even less. It's just
people you know, shouting like mm T memes and stuff.
But then, um, but I would say that in terms
of people at the FED, they're using these simple heuristics
and rules because academic theories there's no help. And what
is when inflation you actually see it. So yes, you're
gonna have to be in the pc data. But is
(11:38):
it going to come down to wage inflation that you
eventually have to see if we're going to see anything sustained?
And at what point does that happen Apart from in
the Bay Area, maybe maybe not so. An interesting thing
about wages is that, uh, when you have moderate inflation,
you know, workers are able to bargain for raises, like
cost of living raises. But then when inflation gets really
high and variable, you know, where it gets up to
(11:59):
like eight or ten percent, but then also bounces around,
workers have a much harder time, you know, anticipating how
much inflation will be and then bargaining to keep up
with that. And so therefore wages you know, often, you know,
at least in the seventies, you know, wages fell behind
during inflation, which doesn't happen in your textbooks but kind
of happens in real life. And so, um, you know,
(12:22):
they're not counting the CPI. Maybe there's some arguments about
that they should be, but the fact is that it's
not wages that's going to show us. It's just you know,
consumer prices, producer prices and stuff like that. So from
your perspective, just to tie it up in a bow,
do you think we've got inflation or not? It's too
early to tell, you know, give it, give it a
few months. Then these things bounce around a lot. It
(12:43):
could be gone next month. This week we also got
a major announcement from Intel, the company's new CEO, Pat Gelsinger,
unveiling an ambitious bid to regain its manufacturing lead by
spending billions of dollars on new factories and creating a
foundry business that will make chips for other companies. Now,
this was not the plan Wall Street had been expecting.
(13:05):
Many thought Intel would go with an ultra light asset
model designing chips and outsourcing, being the capital intensive manufacturing.
We spoke about it with Willy She, Professor Management Practice
at Harvard Business School and an expert in manufacturing and
product development. We started by asking Professor She, just how
difficult this plan is actually going to be for Intel
to execute? Well, uh, twenty billion, that's a start, okay,
(13:32):
I think, uh, you know, it took years for them
to fall behind, Okay, And this is a tough problem.
This is tough technology. Okay. So it's a start, but
clearly Pat Gelsinger has put down a marker and saying
we're climbing back into the ring here and we're gonna
go after leadership again. So you know, I think that's encouraging.
(13:54):
What about where they're doing it in terms of is
it rightly sil SMC and there's an asia full? Is
it right to be focusing so much on what's happening
in Arizona for example? Over in Europe as well well?
I think these shortages that we're seeing, particularly in automotive
but increasingly in other sectors as well, has called attention
(14:16):
to kind of the fragility of supply chains, of vulnerability
the uh kind of excessive dependence on a couple of
suppliers foundries in Taiwan, okay uh. Now, so you see
a lot of that tension in Washington about how we
need to have more of that supply chain domestically. Now,
it's going to take a while to rebuild that. Intel
(14:39):
is proposing to go into the foundry business, I mean,
and they're making a big bet on that. It's going
to take them a while to develop that model. Now.
The good thing is they're starting with two new fabs
that are going to be dedicated to the foundary business.
And what Pat Gelsinger has said is they're going to
do industry to standard tools. They're going to use industry
(15:02):
standard uh I P blocks. Those are the chunks of
circus that you use uh to develop those chips. Right,
So it's going to be more aligned with the model
that other customers who go to the time when these
foundries like t SMC are used to doing. It's not
a model that Intel is used to doing though, So
you know, twenty billion dollars, it's a lot, but even
(15:24):
you know, back in January, t SMC announced it was
going to spend another twenty billion on KPEX. So they're
kind of table sticks for these industries, at these players
to spend a ton of money. When I asked you
is this gonna work, You're like, well, it's a start.
What's it gonna take beyond this initial commitment of an
investment for them to reverse what, as you said, was
(15:44):
like a multi year process of falling behind. Well, you
know what they have to do is they have to
really push the frontiers. Now, what Pat Galsinger has said
is they're going to use they're going to commit to
deep u vy. They've been a little behind on that
right that that, by the way, is one of the
reasons for the twenty billion dollar price tag because when
you look at one of these sml dpub UH machines,
(16:07):
you know, each machine is pushing a hundred fifty million
dollars and in a fab the size that Intel is
going to propose the build, they might need twenty of them. Okay,
So that's like having a fleet to these, you know,
a hundred million plus machines in there. Okay. Now, what
they need to do is they need to get the
(16:28):
processes up and they need to get volume through there, right,
because it's through volume that's how you perfect your processes,
that's how you do your learning, that's how you drive
your costs down. Okay, And t SMC is very good
at that. Of course, they've spent thirty five years developing
that skill. Intel knows how to do that with their microprocessors,
(16:49):
but they've had some troubles with advanced processes, right, so
they'll they'll they'll have to work at it, and it's
going to be multiple years. It's not something UH that
will be able to look two quarters, three quarters, or
even two years out and say, Okay, they're going to
solve this problem. It's going to be a steady progress
(17:10):
and they're just gonna have to take risks and they're
gonna have to make investments. They're gonna have to get
customers to buy that output h so that they can
keep driving volume and learning and almost that exact thing
that it's not going to be solved in a couple
of quarters, that they're going to have to keep on reinvesting.
Can we said exactly the same for the shortage in
chips that we're seeing at the moment, and just take
(17:32):
a listen to what part Els think I had to
say about that. COVID caused everybody to step back a
bit from capacity and build out and the supply chain
the industry, and it induced a radical increase in demand.
So you have supply chains, uh scaling back a bit
and demands scaling up radically. Wow, and now we're in
a position that there's a meaningful shortage and you know
(17:54):
it's going to be a couple of years until that's
fully resolved. Will you have to go on on that
global shortage that we're seeing at the moment and how
that gets resolved? Yeah, Well, there are a lot of
interesting fascests to the shortage. Okay, and we're feeling in
an automotive. But I will remind everybody that you know,
(18:14):
at the beginning of UH, the outlook for the global
automobile market, including China in particular, was slightly down from
UH from two thousand nineteen. And that was before COVID hit, right,
So then COVID hit and sales fall off a cliff UH.
And so if you're an automaker, then you're saying, how
I'm gonna survive, How I'm gonna I'm gonna shepherd cash
(18:36):
and so on, So they pulled back orders. Now, with
the pandemic, did IS created huge demand for things like
note but computers, flat panel displays, consumer electronics, new gaming consoles.
We had new generations that gaming consoles come out. Okay,
and you know, to the extent of the capacity is fungible.
They soaked up a lot of the capacity. So meanwhile,
(18:57):
the auto guys and this started to become visible late
last fall November, it was already starting to become visible.
The lead times for getting your orders into the boundaries
started extending out Now, the US worse by sanctioning Huawei
and sanctioning s M I C so Huawei put in
(19:17):
a lot of orders for chips last summer in Taiwan
and that absorbed a lot of capacity, right, So it
was it was a combination of factors. Now to get
through this uh shortfall, Now, it's gonna take. It's gonna
take a good six months a year before we really
start seeing a lot of relief on that. And we
(19:43):
ended with yet another big tech hearing on Capitol HILP three,
prominent Silicon Valley c e O s Soon Arch Google,
Mark Zuckerberg of Facebook, and Jack Dorsey of Twitter facing
questions from House lawmakers over disinformation on their platform, particularly
in reference to the January insurrection at the Capitol. What's
which was organized online? We've gotten used to these hearings
(20:03):
and they often are not very enlightening or market moving.
But we caught up with Srida Ramaswami, a former senior
VP at Google where he ran their ad business for
fifteen years, to ask whether they will ever be a
meaningful policy turn to open these companies business models and
the threat. Well, I think we need to start with
just having more information part of what is disappointing about
(20:26):
today and honestly about the last three months is we
don't know that much more than what we knew on
January six. Think about it, Joe, if January six had
been an airline accident, the NDSB would have been all
over the place. We would have actual data about what
actually happened. Um while with the tech platforms, we don't
really know how good they were at, say to moving
(20:47):
even illegal content. We're not even talking about things like misinformation.
So I just think we need to be much better educated.
And and you know, our government needs to set up
something like the NDSP to oversee this stuff. If he
can receive airlines, why not orse social media? Only then
will we have meaningful change. Otherwise, you know, we sort
of have this political theater that doesn't really go anywhere importantly.
(21:10):
You say that, of course, as you were overseeing Google's
advertising and commerce products that ended up being a hundred
billion dollars revenue run rate when you were there, Street
and I'm interested you've gone on too, of course, be
a CEO of an add free personal consumer facing search service. Instead.
Your view therefore on on really the regulatory impact of
(21:32):
what you hope to come out of this. Do you
want to see more regulations, just saying, as an oversight boarder,
is there something more deeply that needs to be changed
within section? I'm just saying that oversight more information will
just let us be a lot more educated about what
kind of changes to things like section to thirty are
(21:53):
reasonable to make. I've argued, for example, that we need
to separate things. Downs need to separate out what I
call organic content, content that say, you are Joe share
with your followers on Twitter, versus stuff that's algorithmically amplified
by the platforms. The latter, in my opinion, should be
held to a much higher standard. And the one thing
(22:14):
that they never talked about is all platforms have the
option of delaying content if they're unsure about its nature. Um,
so these are things that they could be doing. Of course,
what I'm doing is completely you know, different from any
of the business models that are out there. Uh nivas
An adds free private search engine because we think having
you as both to use it on customers will just
(22:35):
lead to a better experience over time. So you know,
on that you talk about the distinction that you established
between more organic sharing and sort of more algorithmic things
designed to go viral. I mean people post a lot
of nonsense on different social media platforms, and they often
don't do it as part of some coordinated campaign. They
(22:56):
just tell their friends some misinformation that they heard about
vaccine means or something that they heard about voting machines
or something like That doesn't mean that they like got
it from on top or someone was telling them to
do that. Um. Should the should the platforms think about
throttling that which is just me talking to you excepted um,
(23:16):
you know, completely made up? It's one of the tools
that they have to consider. If it is you know,
our proverbial crazy drunken uncle talking to do office friends,
you know, maybe nothing happens to that. If on the
other hand, if it is a powerful politician with a
megaphone of fifty million people, their words are just completely different.
I think holding that to a higher standard is fine.
(23:39):
Guidelines from Congress, guidelines from regulators are going to help.
But I think we do need to understand that those
two just are not the same. What you and I
might say in the confines of our home to a
couple of friends is completely different from what somebody says
to a hundred million people. Should know how far can
AI go at the moment as it stands, because there's
(23:59):
been in full lot. You hair a lot from Facebook
as the amount of people they have working on this
human eyeballs, but also how they want artificial intelligence to
really be able to pick out the most damaging misinformation,
take it offline as quickly as possible. How good is
the AI getting how quickly can it be enforced? You
think I've dealt with some of the hardest problem at
(24:20):
scale when I was running ADS. AI is not a
magic answer to all solutions, UM, and you run into
the kind of constraints that Joe is talking about, which is,
if you're super accurate at identifying what you think is illegal,
you're going to sweep up a bunch of pretty regular
content along with it. This is why I'm saying we
need more data. We need to know things like if
(24:42):
illegal content is shared on such and such a platform,
they can act on it within two minutes. Are they're
going to delay it for twenty minutes while actual people
look at this. UM At the end of the day,
we need to care about more than effort. We need
to have a shared understanding of the actual thing things
that happen, of actual results. So in my mind. Just
(25:03):
saying we have ten thousand people or twenty thousand people
doesn't let you or me know are they really good
at solving this problem? Yeah, it's part of the answer,
but AI has more constraints than people really want to
talk about. As an investor, thinking about the competitive landscape
for tech, how do you think about the difficulty in
funding companies that theoretically compete with these uh gigantic firms.
(25:27):
How much does that enter into your competition and do
you think that there is a competitive marketplace for the
sort of consumer or other Internet things that that interests you.
It's pretty hard. I think you know, investors understand that
going head to head against very large companies is very problematic.
(25:48):
On the other hand, enterprise investing, that is a there
is a lot of the going on fintech investing, there's
a lot of the going on. In fact, I would
say that NIVA is one of the few pure consumer
more companies that have been funded by vcs in the
recent past um and that's because you know, there is fear.
But going head to head against brilliant dollar companies is
(26:10):
not a smart way to make money. So with your
CEO hat on, your co found a hat on of
NEVA and and being in that competitive space to a
certain degree, Siata, what is the data that you say
you want to see? You say you need more data
to really understand what sort of a job, how good
a job that is they're doing. What data could you
get from these companies? Well, when I when I speak
(26:31):
about social media and data, I'm speaking about it more
as what is the public need or what is the
Congress need in order to reach better decisions? Um? And
to me, as I said, it's it's things like how
quickly subjectionable content brought down, how much of that content
is subject to review before cluse. What's the impact of,
as I said, algorithmic amplification on how much objectionable content
(26:55):
gets shared? Um? Within the context of NEIVA, we solve
a different problem with the different set of principles. UM.
So you know, this kind of data is less of
in the context of my being the CEO of Shouldn't.
But on that note, you say, would they have gladly
give that data? Do you think how would they measure
(27:16):
price to be able to measure those sorts of bits
of data that you're saying that Facebook could give to
the body politics, every team that works on objectionable content
has to grade itself internally in terms of how good
a job they're doing. The metrics that I am talking about,
how quickly subjectionable content identify, how much of it is
(27:40):
sent for human review, how are the human raiders? These
are all things. Yes, the methodology will be different from
company to company, but they have the data, um. But
there needs to be an amount of pressure to have
this data come out in the open so we can
actually make better decisions about what's like, what's the long
(28:00):
term of Section to thirty, what provision to be changed?
How would a job really are we doing? And is
that something that you think, um, the industry should be
more forthcoming with unilaterally on that or is that something
that a regulator or perhaps a legal act could could Enfortunately,
we could actually as you say, good to answer to
see these questions given the past forty years we've been
(28:22):
talking about this since the election, right, Um, I would
actually say that it's going to take some kind of
an act of regulation, perhaps the FTC stepping up to
do the job. I think it's going to require a
certain amount of pressure before more of this information is
is available freely. And that's it for what you missed
(28:46):
this week. Do you like the show and make sure
to subscribe and rate us. An Apple podcast or wherever
you listen to podcast, you can catch our show every
weekday from three thirty to five pm on Bloomberg TV
and from four to five pm on Twitter. Thanks are
listening and have a great week. M H.
Popular Podcasts
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
True Crime Tonight
If you eat, sleep, and breathe true crime, TRUE CRIME TONIGHT is serving up your nightly fix. Five nights a week, KT STUDIOS & iHEART RADIO invite listeners to pull up a seat for an unfiltered look at the biggest cases making headlines, celebrity scandals, and the trials everyone is watching. With a mix of expert analysis, hot takes, and listener call-ins, TRUE CRIME TONIGHT goes beyond the headlines to uncover the twists, turns, and unanswered questions that keep us all obsessed—because, at TRUE CRIME TONIGHT, there’s a seat for everyone. Whether breaking down crime scene forensics, scrutinizing serial killers, or debating the most binge-worthy true crime docs, True Crime Tonight is the fresh, fast-paced, and slightly addictive home for true crime lovers.
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