April 23, 2025 • 42 mins
Today, we're diving deep into the world of commodities with Calculo Capital, a short-term quantitative hedge fund based in Denmark. Our guests, Philip Engel Carlsson and Ole Hansen, share their insights on how they navigate the fast-paced commodity markets using machine learning techniques. We explore how their focus on liquid markets and short holding periods allows them to capitalize on price movements driven by supply and demand imbalances. The conversation also touches on the current trends in commodities, like the rising demand for precious metals and the impact of geopolitical events. Join us as we uncover the strategies behind Calculo's approach and what sets them apart in the investment landscape.
-----
50 YEARS OF TREND FOLLOWING BOOK AND BEHIND-THE-SCENES VIDEO FOR ACCREDITED INVESTORS - CLICK HERE
-----
Follow Niels on Twitter, LinkedIn, YouTube or via the TTU website.
IT’s TRUE ? – most CIO’s read 50+ books each year – get your FREE copy of the Ultimate Guide to the Best Investment Books ever written here.
And you can get a free copy of my latest book “Ten Reasons to Add Trend Following to Your Portfolio” here.
Learn more about the Trend Barometer here.
Send your questions to info@toptradersunplugged.com
And please share this episode with a like-minded friend and leave an honest Rating & Review on iTunes or Spotify so more people can discover the podcast.
Follow Moritz on Linkedin.
Follow Ole on LinkedIn.
Follow Philip on LinkedIn.
Episode TimeStamps:
02:08 - Introduction to Calculo Capital
03:53 - The state of today's commodity markets
10:35 - Commodity performance numbers
11:34 - What is driving the increased performance in the precious metal markets?
15:35 - How the commodity performance impacts the way Calculo Capital is trading
16:48 - Why do Calculo Capital only focus on commodity...
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
(00:00):
Foreign.
And this is where the matchinghappens basically because when we
take this new data put intoour database, it finds a match that
is as close to something ithas seen before.
And as this is historictraits, we do know the outcome of
(00:23):
these particular trades.
C1, C2 so this actually givesus like an indicator on when we have
seen these patterns before.
What happened to thisparticular trade.
Imagine spending an hour withthe world's greatest traders.
Imagine learning from theirexperiences, their successes and
(00:44):
their failures.
Imagine no more welcome to TopTraders Unplugged, the place where
you can learn from the besthedge fund managers in the world
so you can take your managerdue diligence or investment career
to the next level.
Before we begin today'sconversation, remember to keep two
things in all the discussionwe'll have about investment performance
(01:04):
is about the past and pastperformance does not guarantee or
even infer anything aboutfuture performance.
Also understand that there's asignificant risk of financial loss
with all investment strategiesand you need to request and understand
the specific risks from theinvestment manager about their product
before you make investment decisions.
Here's your host, veteranhedge fund manager Niels Kostrup
(01:27):
Larson.
Welcome to another episode inthe Open Interest series on Top Traders
Unplugged, hosted by Moritz Seibert.
In life as well as in trading,maintaining a spirit of curiosity
and open mindedness is key andthis is precisely what the Open Interest
(01:47):
series is all about.
Join Moritz as he engages incandid conversations with seasoned
professionals from around theglobe to uncover their insights,
successes and failures,offering you a unique perspective
on the investment landscape.
So with no further ado, pleaseenjoy the conversation.
(02:08):
Hello and welcome to episodenumber 15 of the Open Interest series
on Top Traders Unplugged.
My name is Murit Siebert.
You know I love finding thesespecialist niche systematic trading
funds and today I'll bespeaking with Kalkulo Capital, a
short term quantitativecommodities hedge fund based in Copenhagen.
Calculo was founded by PhilipEngel Carlson in 2018 after he sold
(02:33):
a power trading company, whichhe founded in 2008 to a competitive
firm.
Prior to getting involved inthe European power markets, Philipp
worked for Saxo Bank'scommodities desk as their Global
Product Manager for futuresand options.
Joining Philip is hiscolleague Ola Hansen, who has been
a member of Kalkulo's board ofdirectors since 2022.
(02:53):
Ole works for Saxo bank inCopenhagen and I reckon some of you
will already be familiar withhim through the widely published
Saxo Weekly Commodity Update,which covers the most important developments
in the global Commoditymarkets as well as through Ole's
appearances on other podcasts.
You know, I wanted to getKalkulo on the Open Interest podcast
for several reasons.
(03:14):
One, there aren't that manyhedge funds around in Denmark to
begin with, at least not to my knowledge.
Second, because they are asystematic cta which instead of trading
a broad set of markets,concentrates on commodities only.
Third, because they are shortterm with holding periods of one
to ten days per trade.
And lastly, because they usemachine learning techniques to generate
(03:37):
trading signals to enter andexit their trades.
So with that as background, Ithink we'll have more than enough
to speak about.
So let me get started bysaying hello to Filip and Ola.
Welcome to both of you.
It's great to have you here.
Thank you very much for having us.
Thank you.
You're more than welcome.
Ola, let's start with you.
You have this detailed viewand this detailed lens into the commodity
(04:00):
markets.
And I always like hearing youranalysis and outlook for the major
markets.
Can you give us a briefsummary of how commodities have performed
here today and everythingthat's been catching your eyes?
Oh yeah, absolutely.
And first of all, it is mostcertainly a very interesting times
across markets, most certainlyalso in commodities.
(04:20):
And we've seen that now for,for the past couple of years where
the, where the interest has, has.
We've seen rising interestamong our clients in Saxo bank and,
and because simply basicallythere's, there's, there's a lot of
different movements going onand what we saw last year was really
quite a diversified marketwhere we had some, some outliers,
(04:41):
we had some like cocoa reallyjumping extremely to the extreme.
And then at the other end wehad some natural gas slumping in
in the US So far this year wereally off to a very strong start.
The, we're seeing all sectorsfooting in some, some pretty strong
gains here.
Well, they are very strongconsidering we only watch six, seven
weeks into the, into the new year.
(05:03):
I watch something like theBloomberg Commodity Index which,
which is broadly exposed toagricultural commodity, metals and
energy and it's up nearly 10%.
And that's most certainly abetter result than what has been
achieved in, in the stockmarket so far this, this year.
And if I should pick out some,some themes.
Well there are, there, thereare many and one, one occurring theme
(05:26):
is, is, is obviously what,what, how the market addresses what's
currently happening in the USwhere we, where we really have some
significant changing signalscoming from the new administration.
The market is, is trying todeal with this as Le at one point
said there are decades wherenothing happens.
(05:47):
And in the weeks where decadeshappen and it almost feels like the
last few weeks has been insomething like that because there
really is a lot going on.
But I think we all watch theprecious metals market where we are
seeing very strong gains notonly last year, but also this year
in precious metals.
And precious metals is whyshould you buy something?
(06:08):
What I think Mr.
Buffett called it a dead asset.
It doesn't yield anything, itdoesn't give you any interest, no
coupon, it costs money tostore and you're constantly worried
someone's going to come andsteal it from you.
But nevertheless, we've seenstrong demand for precious metals
and that really is just anexpression of a world that is not
(06:29):
in balance.
We have an uncertain world andthat's on several fronts.
And that is the reason whywe're seeing investors moving into
precious metals.
Not only central banks whohave been buying for the past three
years in order to diversifytheir holdings away from US dollar
based assets, but also privateindividuals worry about what's seeing,
what they're seeing in theworld right now.
(06:51):
And not only the geopoliticalrisk, but also I think the fiscal
debt worries.
That basically comes with therising debt burdens we're seeing
around the world.
And not only, well, I'll sayalmost, not least in the US right
now, but I think we have along term bullish outlook for commodities.
(07:14):
And there are several keythemes that I think will drive this.
We're seeing thedeglobalization where production
is being moved closer to home.
That unlikely.
That will require quite a lotof commodities.
It will also potentially driveup prices for some.
We have the increased spendingon defense coming through in Europe
now in a big way that willrequire quite a lot of of anything
(07:39):
from steel to rare earthminerals to the more technical side
of things.
We have the wholedecarbonization which is ongoing.
Even though perhaps Trump istrying to put a span in the works.
Then the outlook for futuredemand towards clean energy is ongoing.
And maybe we just call it theenergy transition because the energy
(08:02):
transition is also a questionof we're moving towards increased
use of power.
And it's not only towardselectrical vehicles, it's also towards
data centers, it's towardscooling in countries where climate
change means higher temperatures.
And that leads back to metalsthat will support that process.
(08:24):
Copper and aluminum obviouslysprings to mind.
Lithium, cobalt and nickel arealso there and we see even sea silver.
And then the de dollarizationI mentioned with central banks moving
away from from the, from thedollar that's supporting gold.
We have the disc fiscal anddebt stability risks.
And then finally we've got demographics.
We are basically an agingpopulation in large parts of the
(08:47):
world where with a lot ofmoney but also increasing the pressure
on the younger generation tolook after us.
But also the, the money thatwe have saved in over decades that's
that will be spent and that'salso adding to, to consumption.
And then also we still haveurbanization undergoing especially
in emerging markets that willalso underpin the demand.
And into all this there aresome concerns about supply.
(09:10):
We are seeing that already onthe mining sector where, where some
of the these projects that cantake many many years.
And while we try to, to, to tomake smart decisions at Kakulu on
a, on a one week basis and in,in at my work in probably slightly
longer than that, then themining companies they have to look
10 to 12 years out and try togauge what the price would be out
(09:32):
there.
And that's really, that'sreally holding back some some investment
idea investments and thatpotentially could lead us to this,
to this potential shortfall inin over time in some metals.
So that's what we're focusing on.
But just right now Coco was abig one last year.
This year it has to be, wehave to say it's coffee.
If you look at the soft simplybecause these are exposed to production
(09:55):
or to production fromrelatively small geographic geographical
areas making them much moreexposed to any weather developments.
We're seeing that West Africawith cocoa.
We're seeing that now inBrazil for coffee.
So that is really moving up.
And then we talked aboutbefore we started natural gas is,
is actually on the move andnow we're up 30% this year.
It's after having seen a largedrop last year and I think that's
(10:18):
probably one of the ones whereI have a longer term quite a bullish
view because us US exports ofgas is rising.
US demand for gas as well isalso rising and that will over time
underpin prices.
So that was just not brief.
But there's a lot going on and.
Europe's demand for LNG is rising.
(10:38):
So to put some numbers on it,Net Gas Henry Hub Net gas is up about
30% this year year to date.
Most of that came in the lastcouple of days given a colder weather
outlook for the States.
Coincidentally today is the20th of February.
We got kicked out of a shortHenry Hub position which we've had
on for more than two years.
(10:58):
It's been a great trade for usNatural Gas prices in the US have
been falling quite a bit inrecent years and that rally over
the past couple of days hasput us out of the position.
And like you say, cocoa isdown about 12%, coffee is up 30%
this year, gold is up 15% thisyear already we're not even 60 days
(11:22):
into the new year and gold isup 15%.
I think you've mentioned theindex is up about 10%.
So quite a solid start for commodities.
What do you think Ola?
Copper.
I always think Dr.
Copper is interesting from amacroeconomic perspective as well.
It didn't perform very well inrecent months, but this year already
(11:46):
I think it's up 15 or even abit more than 15%.
What's driving that?
Well, several things I thinkfirst of all foremost the fact that
China hasn't been slapped withterrorists just yet.
That's probably one thing.
But also I think the generalthreat of terrorists has created
(12:06):
quite a few shenanigans in theNew York Comex markets.
And the copper is obviously one.
High grade copper is obviouslyone of those.
So we're seeing right now thatfutures prices for gold, silver,
copper, platinum, they haveall been elevated relatively to international
spot prices, especially thosewe see in London.
And that is due to the worriesthat a lot of these metals has to
(12:30):
be imported.
And suddenly if you have topay in tariffs to get the metals
into the US that obviouslyupsets your whole calculation if
you're short hedging againstphysical positions elsewhere.
And that's where we often seethe futures market being used because
with depth of liquidity andround the club trading it is the
favorite place go to placewhere investors and physical traders
(12:53):
they offset their risk in thefutures market.
But obviously short positiondoesn't work if you actually need
delivery.
And there's terrorists and wesaw that actually just last Friday.
If we look at the price basedon London metals, at one point the
New York spiked to $1,000premium over new York.
That has since reverted a bit lower.
(13:14):
But it just highlights a lotof uncertainty.
We've seen massive amount ofsilver and gold being shipped to
New York in the past couple ofweeks simply because if you have
access to physicals and youcan sell the future against it, you
are earning a handsome return.
And the question willobviously be what happens when these
spreads normalizes?
Then there's a lot of physicalgold sitting in the US with no end
(13:38):
user.
Will that then go back intothe market or will we continue to
see selling hedging hedgeselling against it which potentially
could have a small Negativeimpact on prices, but a lot of things
going on in that market.
But copper, I think justgenerally it is a market where the
demand in China has held upphenomenally well, even though the
(13:59):
construction sector has goneinto reverse.
And as we highlighted the.
Right now we're seeing thatwith the data centers and the extra
amount for electricity, copperis not only an EV story and a data
center story, it's also beingable to transmit all this power that's
going to be needed.
That's why a company likeSiemens Energy is up what 300% in
(14:22):
the last year because that'sreally probably where a lot of the
gains are being made.
Bit like the Nvidia that it'spotentially in like the gold digging
days.
It wasn't a gold miner or golddigger who made the money, it was
the guy who supplied the shovels.
And that's what we're seeingright now also in the power space
(14:42):
that those that are buildingthese transcendors and making transmission
possible there, they're very busy.
I picked up that the cash andcarry bases on gold is something
like 10% annualized on COMEX.
Yep.
And it, it's probably evenstronger right now in silver because
there's no last resort lendingof last resort in the, in the silver
(15:05):
market because if you are, ifyou have a, if you're short gold
in London market you can go tothe bank of England and, and lease
gold from, from them.
There's not no such holdingsof silver.
So that's why silver has beenbeen the trick hard hit also or squeezed
also because a lot of thatcomes from Mexico and we're still
unclear about.
And there's probably morehigher risk that tariffs would be
(15:28):
added to silver than to goldbeing a monetary, you almost say
monetary currency.
So Philip, with all thesesuper interesting things and the
knowledge that Olay has justoutlined here, does any of that impact
what Kalkulu is doing on a dayto day basis and how you trade?
Yeah.
Moritz, first of all justthank you for having us here at your
(15:51):
show.
As a regular listener myself,I really enjoy being here.
So thank you so much.
And of course the impact oncommodities worldwide do of course
impact Calculo, but as youmentioned we are focused on the very
short end.
(16:11):
So we have a holding period ofas you said like one to 10 days.
It's more like two to fivedays actually.
And this allow us to captureon these imbalances as Ole mentions
in his commodity research here.
And what we do is reallycapturing the alpha and then ride
(16:34):
this alpha movement and thenwe use the AI to navigate our exits
even better.
So for us it's capturing thealpha moves and then use AI to get
an exit on our positions.
Great, let's go step by step.
I mean, one of the firstthings that jumped to my mind that
(16:55):
I'd like to ask you is why didyou end up focusing only on the commodity
markets?
I mean you could get morediversification by trading additional
markets such as bonds andequities and currencies, some of
the financials, but you don'tdo that for a reason.
It's a feature and not a bug.
You only trade the commodities.
(17:15):
Why is that?
That's true.
First of all, my heart is for commodities.
But that is not the onlyreason for this limitation.
In the, in the space we dotrade in the supply demand constraints
on these particularinstruments is what we actually centered
(17:36):
our strategy about.
So the, for instance whenwe're seeing a supply constraint
due to a weather patternregional event that creates fear
into the market, this fearsparks price movements.
And these movements isactually what we want to capture
in our trend following model.
(17:57):
This is not directly somethingyou can do in the equity market or
in the interest rate market,but it's really comes very natural
into the supply demandconstraints on the commodities.
Why this strategy worksparticularly well in these environments.
Do you focus on a subset ofthe commodities?
Because you just mentionedweather dependencies.
(18:20):
For instance, gold would notbe weather dependent, but you know,
coffee and cocoa are.
Or are you trading across theboard all the commodity markets or
everything that is liquidenough for your definition?
You kind of answered thequestion yourself.
We only do trade the liquid stuff.
So we are in the four majorasset classes or groups you can call
(18:43):
like metals, agricultural,softs and energies.
And we focus on where theliquidity is the highest good because
that also supplements ourstrategy quite well.
We don't trade some of thesmaller commodities, even though
I know that some of the otherCTA funds you as well, I think participates
(19:05):
in the orange juice market for instance.
That has been a quite a goodtrade for a lot of the CTA funds.
But we do tend to focus on theliquid stuff where we can also grow
our fund quite, quite a bit.
So yeah, so we are the majorcommodities basically, right?
I mean clearly you're tradingway more short term than, than we
(19:28):
do with two to five days, Ithink you've mentioned.
I mean the orange juiceposition for us has been going on
for way more than two years.
So you know, we don't reallyneed to tap into the liquidity of
that market on a daily basis.
But with your tradingfrequency you're absolutely much
more sensitive to that.
Then coming back to the supplydemand imbalances, I mean it is represented
(19:57):
by price at the end of the day.
What is your definition of a,you know, imbalance and how do you
spot that?
Like what triggers this?
Basically we use only pricepatterns as our entry.
We do lean towards veryclassical CTA mythology because I'm
(20:19):
a very strong believer of a.
Of the CTA fund and the corearound the CTA funds modern day trading.
We do need to modify the veryclassical the way that the CTA evolved
and started.
So while we use like a veryopportunistic entry signals, very
(20:41):
classical CTA style breakoutspredominantly we only use the entry
as, as, as the first part ofthe trade.
The real advanced stuffactually comes in after we have the
trade on the books.
And that is our risk management.
(21:01):
One of our core fundamentalsis that we do want to deliver a very
low volatility access to thecommodity markets.
So why we do love commodities,Commodities can be volatile and,
and through the use of AI wecan sort out a lot of the volatility
and create a more linearyreturn profile.
(21:25):
Right.
So breakout to the upside orto the downside that would trigger
the position.
You've mentioned Philip, acouple of minutes ago, weather dependency.
So you're not using weatherforecasts as an input in your trading
signals.
It is really price based.
And price pattern based Ithink is what I understood.
So search sequence of priceevents materializing or the price
(21:45):
making a new high.
And like a trend followingtrader you would then get long once
that, you know, high has been made.
That is true in regards to theweather pattern.
It really comes into the price.
Let's say we have a hurricanethreatening the oil rigs, gas rigs
in the, in the Gulf of Mexicoor Gulf of America, what we call
it these days.
(22:06):
That is actually what we arelooking at because this threat from
the hurricane will trigger aprice movement that we can benefit
from.
And after this initial riskpremium being put into the market,
the market will look at thehurricane development, the speed
(22:27):
the, the classification andwhen it goes landfall and at what
strengths it goes landfall at.
And this period when weactually in this limbo, it gives
us a lot of underlying signalsthat we can use in our machine learning
regard to the momentum,candlestick recognition patterns
and so forth that we use toclassify and use our past history
(22:51):
on the TRA and the trace wehave already made to see if we can
predict the upcoming movementof these particular commodities.
So quite often we doparticipate in these movements where
the market stabilizes a bitand trade sideways.
And this is often enough forus to exit the position capturing
(23:13):
the alpha and then look for anew signal into the same trend if
it continues up.
For instance, is this a fullysystematic process?
Like do your systems pick upthat there is a hurricane potentially
materializing say in Septemberor October and it automatically filters
through to your machinelearning environment?
(23:34):
Or is, is it a discretionaryinput where you need, where you need
to nut your system and tellthe system hey, there's a hurricane
potentially materializing.
You should have a look at this data.
No, it's all price recognition patterns.
So we only use prices input.
There's no manual input inregards to selecting signals.
(23:54):
It all be handled by ourproprietary trading platform.
So we don't use any, anyweather patterns or any manual or
writing in the signalselection process.
Got it, Got it.
And just coming back to theentry signals, you're not forcing
your system to be long andshort at all times.
(24:15):
You can be flat as well, right?
If there's nothing going on,then you don't have a position in
whatever market doesn'tinterest the system.
Exactly, yeah, spot on.
So basically as we only wantto participate in in real alpha periods,
we can be flat and we can beon a partly long allocation or partly
(24:36):
short allocation or combination.
So we don't have any mandatoryallocation towards individual markets
or sectors or groups or directional.
We can be in anywhere themarket is.
Basically let's then move astep forward.
Now you have a position, it'sbeen triggered, let's say a long
position has been triggered bya breakout to the upside and the
(24:58):
expectation is now that you'rein that trade for two to five days.
Is the logic a trend followinglogic in the sense that you would
hold onto the position untilan exit is elected or hit by price,
which is on a reversal you'renever getting out on the top.
Or you have more differentexit techniques where you're taking
(25:21):
profits.
Basically you can divide ourmythology into two groups where we
have the, the systematicapproach, where we do very classical
cta, we enter into a trade, weput in our stops, we use trailing
stops as well.
But this is only one part ofthe trading strategy.
(25:44):
As such, because our majorityof our trades actually is not held
onto the trading stops, we usethe machine learning exits which
takes out, I would say 80, 75,80% of our trades is handled via
the machine learning exits.
(26:05):
So basically what happens isthat on a daily basis we Collect
a lot of data concerning theindividual trade we have in our books.
This could be candlestickpattern recognition, momentum, angle
on momentum and otherindicators as like that is part of
explaining the currentmovement on or the current short
(26:29):
term trend.
We then use this data patternthat we have collected on the individual
trade on the day we go intoour massive massive database where
we can see if we can find asimilar pattern as we have collected
on the day for this particular trade.
(26:50):
And this is where the matchinghappens basically because when we
take this new data put intoour database, it finds a match that
is as close to something ithas seen before.
And as this is historictrades, we do know the outcome of
these particular trades.
T +1, T +2.
So this actually gives us likean indicator on when we have seen
(27:15):
these patterns before whathappened to this particular trig.
And this is basically what we use.
And then we classify them askeep or eliminate.
And if it's eliminate then wetake them off the books and look
for new entries.
Basically.
Are you streaming tick data toyour systems or is it batched into
(27:35):
15 or 30 minute bars?
Or I assume that you're usingmore frequent data than a daily open,
high, low close for what it isthat you do.
We use various time frames,but when we do it on the close in
regards to the machinelearning, we actually put emphasis
into the daily bars and thedaily data.
(27:56):
Because as I mentioned before,we do believe in the CTA mythology,
We do like trends and all that.
So we just classify the trendsdifferently than a traditional CTA.
So while classical CTAs are intrends for the long term, for the
long time, in the long run wetry to capture bits and pieces of
(28:18):
these existing trends, tryingto vet out any counter moves along
the way.
Doing that allows us to maybenot benefit from the entire trend,
but we do capture part of thetrend and we vet out the volatility
allowing us to create a muchsmoother ride than than the longer
(28:41):
term trend followers.
And all the entries entrysignals are pure trend following
signals.
Or do you also combine meanreversion and other type of methods
and strategies into the mix?
We use trend followingstrategies to enter.
The markets trend only.
Okay, so.
So let's have a little bit ofa chat about machine learning and
(29:02):
we're not using it.
I'm definitely not the bestperson to speak about it because
I don't have much experiencewith it.
But how does it work?
Like how do you prevent themachine to learn from the noise which
is a prevalent factor in the markets?
I mean there's also a lot ofsilly things that a machine can learn.
You want it to learn the good things.
(29:23):
I sometimes joke that a clevermachine will probably learn to buy
the breakout and become atrend following trader long term.
But I may be wrong.
So how do you make sure youunderstand what's happening in the
black box?
Or is there even a need foryou to understand?
Or could you just say well wedon't care, we just want to make
money, it's fine as well.
That's a very good question actually.
(29:45):
And keep in mind that we donot use machine learning techniques
for the entry and thatactually answers your concern because
when you use machine learningfor the entry signals then you actually
put great emphasis intoguessing where the market is is moving.
(30:06):
And I'm not a strong believerin in that part actually.
So I do believe that trendshave strength.
So running along a trend is inmy opinion a very strong part of
this environment that we're in.
We only use machine learningas a risk of mythology and this actually
(30:31):
changed the picture quite abit because if you use machine learning
as a risk of mythology,basically let's say if we go back
to the hurricane example weused before.
Market has risen quite rapidlydue to a hurricane threat.
After the first initial riskpremium being put into the market,
the market stabilizes.
(30:53):
Everybody's watching theweather reports the path of the hurricane
and all that stuff.
So the market stabilizes andtrades sideways.
Maybe even some of the traderstake profit on the initial move.
So it might maybe even liketrace back a bit.
As a short term trader oursystem reads the short term trends
(31:14):
towards the long trend.
So these movements and theactions around this stabilizing period
is enough for us to get validdata into getting signals on how
to take when to take off this position.
So we log in the first part ofthis risk premium, take the profits
(31:37):
and then look for a new signal.
So when the market eithercollapses because the hurricane didn't
hit or hit the the oil rigsand cause lot of dis destruction
and the trend continues, wecan actually re enter into the same
trend that we captured thefirst alpha move on and get the second
(31:58):
leg of this particular trade.
So we actually it only uses itas a risk off element which changes
the picture quite a bitcompared to trying to build a fully
AI model that tries to predictthe price targets.
Yeah.
So it's all happening post entry.
(32:20):
Once you have a position on inthe materials Philip that you kindly
shared with me ahead of thecall they mentioned adaptive execution
is the adaptive executionpiece what you mean with getting
out of the position andessentially taking it off off using
machine learning techniques?
Or does it also mean that youhave a more smart, clever way of
(32:40):
executing your trades in themarket using alos given that you
are short term?
You know, I reckon that youhave some sensitivity toward that.
Sure, both.
So we, we both use advancedALOS to get into the market.
But the adaptive stuff is alsovery, very much in meant as the exit
(33:01):
period as well.
We also use like a time of theexits as a parameter as well where
we try to get our exits inliquid periods predominantly.
So of course our stops areexecuted as it happens, but our machine
learning exits actually timedtowards the most liquid time in the
(33:24):
US because remember that thesemachine learning classifications
are not firm predictions thatthe market will collapse or go against
your current position.
It's just a flag saying beaware there's a potential correction
ahead of you.
(33:44):
And this is why we take morein the liquid periods.
Right.
Creating a segue to OLA here.
Ola, I'm sure looks at all thecommodity forward curves, whether
they're in contango orequitation and seasonal and all that.
Does this information triggeranything at Kalkulu or are you maybe
even trading spreads which youknow, you can do in the commodity
(34:06):
markets?
We're not showing spreads as is.
No.
Okay, but Ola, you have allthis information about the forward
curves.
You have this view into thecommodity markets.
But I also know that you're aninvestor in Kalkulo.
What has triggered your choiceto you know, put your money into
a quantitative long shortcommodity strategy as opposed to
(34:29):
one that is fundamentallydriven where you know, we just spoke
about copper and cocoa andcoffee and all the views that you
can have on these markets.
Well, I, I have it as, as, aspart of a general exposure because
I, I like commodities.
I, I've been involved withcommodities for, for more than 20
years.
And, and it's, it's reallywhere I, I, I feel that as I have
(34:51):
an edge but obviously I knownthis the Kakul for many years and
I have been involved for anumber of years.
Probably what attracted meinto an investment originally was
simply the low volatility and returns.
The correlation betweenreturns and the low volatility really
(35:12):
quite appealed to me.
And then there's also simplyto have a different approach instead
of just a long only.
And, and, and when you lookat, when you look back and we look
at some of the performances,it's actually interesting to see
that the, the Bloombergcommodity index that I do that I
do follow obviously on a dailybasis that the performances difference
is not great.
But the road travel to get towhere we are today is massively different.
(35:36):
And that's why I prefer thiskind of strategy because just looking
at the Bloomberg commoditysince in the last five years we've
had three we had several drawdowns.
We had the two biggest one wasa 77 months drawdown between 15 and
20 where it fell 38% then wehad a 33 month drawdown up until
now we still haven't reachedthe top in 2022 but the initial drop
(35:59):
off in this latest correctionwhich we're now trying to reverse
was 22% and that is reallythese kind of drawdowns is, is what
you have to have to accept inorder to get this long term return
which then ultimately end upbeing not being significantly different
to what what Kula has has, haspresented with a, with a much lower
(36:23):
volatility.
So that's really what whatwhat attracted me into it.
And I think also also what Ihave noticed is on, on some of the
months where we've had the, wehaven't had many but I think we're
probably going to get more inthe future where we had to have bad
months for equities we quiteoften find that there's actually
some positive months coming inthrough the commodities.
(36:45):
So also just generallylowering the overall volatility across
your investments.
I mean commodities don't havethe equity drift or anything like
that.
I really think there aremarkets that you should be trading
long and short and when youlook at the BCOM index, well first
off all of these indices, thefirst generation of these indices
(37:07):
they're equity and metalsheavy and then there's all the smart
variations that are morediversified and they distribute liquidity
across the curve more.
But I think COCO for instancebecause it's a relatively small market
is not a BCom member probablyfor the last two years.
They wish they had it in thereit would have made a difference.
But luckily you can say thisyear last year it was the natural
(37:29):
gas which is member that drawdragged it down and cocoa which was
was not involved.
And this year it's, it'sthere's now coffee that is both coffee
and natural gas.
That's that's putting butputting higher.
But yeah no doubt that some ofthe others last year had a very good
year but also meant that therewas a massive re rebalancing in in
the, when the, when theJanuary came because the the percentage
(37:51):
share of or exposure to cocoaout of the total was uh.
Had gone quite extreme becauseof the, the big gains we saw last
year.
So, so yeah, there are, thereare, there are different ways.
I also watch ETFs where theytry to be smart on the role.
It hasn't been.
I don't find them having beenparticularly successful.
They haven't reallyoutperformed the main total return
(38:15):
in the indices.
So even though we are in asituation that you mentioned contanguant
backwardation where wegenerally are starting to see a positive
role yield coming into the markets.
Except.
And the biggest option isobviously where you just have had
your profitable short innatural gas.
Because.
Because the contango structurebasically means this is a market
that even though we're seeingsometimes seeing big rises over time,
(38:38):
you're actually making money.
You have been making moneybeing short simply on the roll from
a low to a high contract every month.
Well that's part of it.
We're not rolling every month.
We're aware of seasonalities.
But yes, for sure we didbenefit from the generally upward
sloped forward curve there.
(38:59):
Philip, I think you wanted tojump in and say something.
Yeah, that's right.
Just to supplement Ola'scomments here.
One of the things that weactually also have seen in the Calculo
fund is that we over the pastseven years almost since inception
has actually delivered equitylike returns at two thirds of the
(39:21):
volatility of the broad stock indices.
This is also just one of theselling points for Kalkulo strategy
being a low volatilitystrategy and non correlated, non
cyclic.
So actually allows a lot ofour investors and we see that a lot
when we, when we meet ourclients that we can actually offer
something different to themthat they can get from other asset
(39:44):
classes.
We are liquid, we're notcorrelated and stuff.
So it's really an add on fromfor many of our clients in this space.
Sure.
One question I'd have PhilipBenjola is you're domiciled in Denmark
and I think you're the onlyfund that I know that is domiciled
in Denmark.
How did that come about?
I mean it is not known as ahedge fund land.
(40:08):
Neither is Germany for sure.
But you know people go toCayman or Luxembourg Island.
You're in Denmark.
Why that?
First of all we are Danish.
Sure, that is one thing.
But yeah you're definitely right.
I think we might have broaderaudience in a different country than
Denmark but sometimes havingthe space on your own is also an
(40:31):
advantage.
So we are probably the onlypure commodity fund in Denmark and
it's our job to educate and toget the word out there.
And anything that's on thehorizon for you guys in terms of
how you're looking to growyour business and develop it, maybe
become more international.
(40:53):
Definitely.
We have explored the optionsof doing like a US setup because
strategies like the Kulostrategy would really thrive in the
US I believe.
So this definitely one of ourthings on the agenda to look abroad
to see if we can actuallyexpand our client network in more
(41:19):
commodity knowledgeable areasas well.
Would you also trade SMAs orare you purely focused on managing
funds?
We'd be looking into both, actually.
We wouldn't rule out SMAs thatwould require a US license for that,
of course, but definitely it'sall in the pipeline.
(41:42):
Excellent guys.
Thanks very much for comingonto the podcast.
I really enjoyed it.
I hope our listeners will findthe conversation valuable and interesting
as well.
As usual, we'll include themost important points of today's
discussion in our show notes.
And also, as usual, youshouldn't hesitate to contact us
if you have any questions.
Our email is infooptradersunplugged.com so thanks again for
(42:06):
listening.
And until next time on TopTraders Unplugged.
Thanks for listening to TopTraders Unplugged.
If you feel you learnedsomething of value from today's episode,
the best way to stay updatedis to go on over to itunes and subscribe
to the show so that you'll besure to get all the new episodes
as they're released.
We have some amazing guestslined up for you, and to ensure our
(42:27):
show continues to grow, pleaseleave us an honest rating and review
in itunes.
It only takes a minute andit's the best way to show us you
love the podcast.
We'll see you next time on TopTraders Unplugged.
Foreign.
And this is where the matchinghappens basically because when we
take this new data put intoour database, it finds a match that
is as close to something ithas seen before.
And as this is historictraits, we do know the outcome of
(00:23):
these particular trades.
C1, C2 so this actually givesus like an indicator on when we have
seen these patterns before.
What happened to thisparticular trade.
Imagine spending an hour withthe world's greatest traders.
Imagine learning from theirexperiences, their successes and
(00:44):
their failures.
Imagine no more welcome to TopTraders Unplugged, the place where
you can learn from the besthedge fund managers in the world
so you can take your managerdue diligence or investment career
to the next level.
Before we begin today'sconversation, remember to keep two
things in all the discussionwe'll have about investment performance
(01:04):
is about the past and pastperformance does not guarantee or
even infer anything aboutfuture performance.
Also understand that there's asignificant risk of financial loss
with all investment strategiesand you need to request and understand
the specific risks from theinvestment manager about their product
before you make investment decisions.
Here's your host, veteranhedge fund manager Niels Kostrup
(01:27):
Larson.
Welcome to another episode inthe Open Interest series on Top Traders
Unplugged, hosted by Moritz Seibert.
In life as well as in trading,maintaining a spirit of curiosity
and open mindedness is key andthis is precisely what the Open Interest
(01:47):
series is all about.
Join Moritz as he engages incandid conversations with seasoned
professionals from around theglobe to uncover their insights,
successes and failures,offering you a unique perspective
on the investment landscape.
So with no further ado, pleaseenjoy the conversation.
(02:08):
Hello and welcome to episodenumber 15 of the Open Interest series
on Top Traders Unplugged.
My name is Murit Siebert.
You know I love finding thesespecialist niche systematic trading
funds and today I'll bespeaking with Kalkulo Capital, a
short term quantitativecommodities hedge fund based in Copenhagen.
Calculo was founded by PhilipEngel Carlson in 2018 after he sold
(02:33):
a power trading company, whichhe founded in 2008 to a competitive
firm.
Prior to getting involved inthe European power markets, Philipp
worked for Saxo Bank'scommodities desk as their Global
Product Manager for futuresand options.
Joining Philip is hiscolleague Ola Hansen, who has been
a member of Kalkulo's board ofdirectors since 2022.
(02:53):
Ole works for Saxo bank inCopenhagen and I reckon some of you
will already be familiar withhim through the widely published
Saxo Weekly Commodity Update,which covers the most important developments
in the global Commoditymarkets as well as through Ole's
appearances on other podcasts.
You know, I wanted to getKalkulo on the Open Interest podcast
for several reasons.
(03:14):
One, there aren't that manyhedge funds around in Denmark to
begin with, at least not to my knowledge.
Second, because they are asystematic cta which instead of trading
a broad set of markets,concentrates on commodities only.
Third, because they are shortterm with holding periods of one
to ten days per trade.
And lastly, because they usemachine learning techniques to generate
(03:37):
trading signals to enter andexit their trades.
So with that as background, Ithink we'll have more than enough
to speak about.
So let me get started bysaying hello to Filip and Ola.
Welcome to both of you.
It's great to have you here.
Thank you very much for having us.
Thank you.
You're more than welcome.
Ola, let's start with you.
You have this detailed viewand this detailed lens into the commodity
(04:00):
markets.
And I always like hearing youranalysis and outlook for the major
markets.
Can you give us a briefsummary of how commodities have performed
here today and everythingthat's been catching your eyes?
Oh yeah, absolutely.
And first of all, it is mostcertainly a very interesting times
across markets, most certainlyalso in commodities.
(04:20):
And we've seen that now for,for the past couple of years where
the, where the interest has, has.
We've seen rising interestamong our clients in Saxo bank and,
and because simply basicallythere's, there's, there's a lot of
different movements going onand what we saw last year was really
quite a diversified marketwhere we had some, some outliers,
(04:41):
we had some like cocoa reallyjumping extremely to the extreme.
And then at the other end wehad some natural gas slumping in
in the US So far this year wereally off to a very strong start.
The, we're seeing all sectorsfooting in some, some pretty strong
gains here.
Well, they are very strongconsidering we only watch six, seven
weeks into the, into the new year.
(05:03):
I watch something like theBloomberg Commodity Index which,
which is broadly exposed toagricultural commodity, metals and
energy and it's up nearly 10%.
And that's most certainly abetter result than what has been
achieved in, in the stockmarket so far this, this year.
And if I should pick out some,some themes.
Well there are, there, thereare many and one, one occurring theme
(05:26):
is, is, is obviously what,what, how the market addresses what's
currently happening in the USwhere we, where we really have some
significant changing signalscoming from the new administration.
The market is, is trying todeal with this as Le at one point
said there are decades wherenothing happens.
(05:47):
And in the weeks where decadeshappen and it almost feels like the
last few weeks has been insomething like that because there
really is a lot going on.
But I think we all watch theprecious metals market where we are
seeing very strong gains notonly last year, but also this year
in precious metals.
And precious metals is whyshould you buy something?
(06:08):
What I think Mr.
Buffett called it a dead asset.
It doesn't yield anything, itdoesn't give you any interest, no
coupon, it costs money tostore and you're constantly worried
someone's going to come andsteal it from you.
But nevertheless, we've seenstrong demand for precious metals
and that really is just anexpression of a world that is not
(06:29):
in balance.
We have an uncertain world andthat's on several fronts.
And that is the reason whywe're seeing investors moving into
precious metals.
Not only central banks whohave been buying for the past three
years in order to diversifytheir holdings away from US dollar
based assets, but also privateindividuals worry about what's seeing,
what they're seeing in theworld right now.
(06:51):
And not only the geopoliticalrisk, but also I think the fiscal
debt worries.
That basically comes with therising debt burdens we're seeing
around the world.
And not only, well, I'll sayalmost, not least in the US right
now, but I think we have along term bullish outlook for commodities.
(07:14):
And there are several keythemes that I think will drive this.
We're seeing thedeglobalization where production
is being moved closer to home.
That unlikely.
That will require quite a lotof commodities.
It will also potentially driveup prices for some.
We have the increased spendingon defense coming through in Europe
now in a big way that willrequire quite a lot of of anything
(07:39):
from steel to rare earthminerals to the more technical side
of things.
We have the wholedecarbonization which is ongoing.
Even though perhaps Trump istrying to put a span in the works.
Then the outlook for futuredemand towards clean energy is ongoing.
And maybe we just call it theenergy transition because the energy
(08:02):
transition is also a questionof we're moving towards increased
use of power.
And it's not only towardselectrical vehicles, it's also towards
data centers, it's towardscooling in countries where climate
change means higher temperatures.
And that leads back to metalsthat will support that process.
(08:24):
Copper and aluminum obviouslysprings to mind.
Lithium, cobalt and nickel arealso there and we see even sea silver.
And then the de dollarizationI mentioned with central banks moving
away from from the, from thedollar that's supporting gold.
We have the disc fiscal anddebt stability risks.
And then finally we've got demographics.
We are basically an agingpopulation in large parts of the
(08:47):
world where with a lot ofmoney but also increasing the pressure
on the younger generation tolook after us.
But also the, the money thatwe have saved in over decades that's
that will be spent and that'salso adding to, to consumption.
And then also we still haveurbanization undergoing especially
in emerging markets that willalso underpin the demand.
And into all this there aresome concerns about supply.
(09:10):
We are seeing that already onthe mining sector where, where some
of the these projects that cantake many many years.
And while we try to, to, to tomake smart decisions at Kakulu on
a, on a one week basis and in,in at my work in probably slightly
longer than that, then themining companies they have to look
10 to 12 years out and try togauge what the price would be out
(09:32):
there.
And that's really, that'sreally holding back some some investment
idea investments and thatpotentially could lead us to this,
to this potential shortfall inin over time in some metals.
So that's what we're focusing on.
But just right now Coco was abig one last year.
This year it has to be, wehave to say it's coffee.
If you look at the soft simplybecause these are exposed to production
(09:55):
or to production fromrelatively small geographic geographical
areas making them much moreexposed to any weather developments.
We're seeing that West Africawith cocoa.
We're seeing that now inBrazil for coffee.
So that is really moving up.
And then we talked aboutbefore we started natural gas is,
is actually on the move andnow we're up 30% this year.
It's after having seen a largedrop last year and I think that's
(10:18):
probably one of the ones whereI have a longer term quite a bullish
view because us US exports ofgas is rising.
US demand for gas as well isalso rising and that will over time
underpin prices.
So that was just not brief.
But there's a lot going on and.
Europe's demand for LNG is rising.
(10:38):
So to put some numbers on it,Net Gas Henry Hub Net gas is up about
30% this year year to date.
Most of that came in the lastcouple of days given a colder weather
outlook for the States.
Coincidentally today is the20th of February.
We got kicked out of a shortHenry Hub position which we've had
on for more than two years.
(10:58):
It's been a great trade for usNatural Gas prices in the US have
been falling quite a bit inrecent years and that rally over
the past couple of days hasput us out of the position.
And like you say, cocoa isdown about 12%, coffee is up 30%
this year, gold is up 15% thisyear already we're not even 60 days
(11:22):
into the new year and gold isup 15%.
I think you've mentioned theindex is up about 10%.
So quite a solid start for commodities.
What do you think Ola?
Copper.
I always think Dr.
Copper is interesting from amacroeconomic perspective as well.
It didn't perform very well inrecent months, but this year already
(11:46):
I think it's up 15 or even abit more than 15%.
What's driving that?
Well, several things I thinkfirst of all foremost the fact that
China hasn't been slapped withterrorists just yet.
That's probably one thing.
But also I think the generalthreat of terrorists has created
(12:06):
quite a few shenanigans in theNew York Comex markets.
And the copper is obviously one.
High grade copper is obviouslyone of those.
So we're seeing right now thatfutures prices for gold, silver,
copper, platinum, they haveall been elevated relatively to international
spot prices, especially thosewe see in London.
And that is due to the worriesthat a lot of these metals has to
(12:30):
be imported.
And suddenly if you have topay in tariffs to get the metals
into the US that obviouslyupsets your whole calculation if
you're short hedging againstphysical positions elsewhere.
And that's where we often seethe futures market being used because
with depth of liquidity andround the club trading it is the
favorite place go to placewhere investors and physical traders
(12:53):
they offset their risk in thefutures market.
But obviously short positiondoesn't work if you actually need
delivery.
And there's terrorists and wesaw that actually just last Friday.
If we look at the price basedon London metals, at one point the
New York spiked to $1,000premium over new York.
That has since reverted a bit lower.
(13:14):
But it just highlights a lotof uncertainty.
We've seen massive amount ofsilver and gold being shipped to
New York in the past couple ofweeks simply because if you have
access to physicals and youcan sell the future against it, you
are earning a handsome return.
And the question willobviously be what happens when these
spreads normalizes?
Then there's a lot of physicalgold sitting in the US with no end
(13:38):
user.
Will that then go back intothe market or will we continue to
see selling hedging hedgeselling against it which potentially
could have a small Negativeimpact on prices, but a lot of things
going on in that market.
But copper, I think justgenerally it is a market where the
demand in China has held upphenomenally well, even though the
(13:59):
construction sector has goneinto reverse.
And as we highlighted the.
Right now we're seeing thatwith the data centers and the extra
amount for electricity, copperis not only an EV story and a data
center story, it's also beingable to transmit all this power that's
going to be needed.
That's why a company likeSiemens Energy is up what 300% in
(14:22):
the last year because that'sreally probably where a lot of the
gains are being made.
Bit like the Nvidia that it'spotentially in like the gold digging
days.
It wasn't a gold miner or golddigger who made the money, it was
the guy who supplied the shovels.
And that's what we're seeingright now also in the power space
(14:42):
that those that are buildingthese transcendors and making transmission
possible there, they're very busy.
I picked up that the cash andcarry bases on gold is something
like 10% annualized on COMEX.
Yep.
And it, it's probably evenstronger right now in silver because
there's no last resort lendingof last resort in the, in the silver
(15:05):
market because if you are, ifyou have a, if you're short gold
in London market you can go tothe bank of England and, and lease
gold from, from them.
There's not no such holdingsof silver.
So that's why silver has beenbeen the trick hard hit also or squeezed
also because a lot of thatcomes from Mexico and we're still
unclear about.
And there's probably morehigher risk that tariffs would be
(15:28):
added to silver than to goldbeing a monetary, you almost say
monetary currency.
So Philip, with all thesesuper interesting things and the
knowledge that Olay has justoutlined here, does any of that impact
what Kalkulu is doing on a dayto day basis and how you trade?
Yeah.
Moritz, first of all justthank you for having us here at your
(15:51):
show.
As a regular listener myself,I really enjoy being here.
So thank you so much.
And of course the impact oncommodities worldwide do of course
impact Calculo, but as youmentioned we are focused on the very
short end.
(16:11):
So we have a holding period ofas you said like one to 10 days.
It's more like two to fivedays actually.
And this allow us to captureon these imbalances as Ole mentions
in his commodity research here.
And what we do is reallycapturing the alpha and then ride
(16:34):
this alpha movement and thenwe use the AI to navigate our exits
even better.
So for us it's capturing thealpha moves and then use AI to get
an exit on our positions.
Great, let's go step by step.
I mean, one of the firstthings that jumped to my mind that
(16:55):
I'd like to ask you is why didyou end up focusing only on the commodity
markets?
I mean you could get morediversification by trading additional
markets such as bonds andequities and currencies, some of
the financials, but you don'tdo that for a reason.
It's a feature and not a bug.
You only trade the commodities.
(17:15):
Why is that?
That's true.
First of all, my heart is for commodities.
But that is not the onlyreason for this limitation.
In the, in the space we dotrade in the supply demand constraints
on these particularinstruments is what we actually centered
(17:36):
our strategy about.
So the, for instance whenwe're seeing a supply constraint
due to a weather patternregional event that creates fear
into the market, this fearsparks price movements.
And these movements isactually what we want to capture
in our trend following model.
(17:57):
This is not directly somethingyou can do in the equity market or
in the interest rate market,but it's really comes very natural
into the supply demandconstraints on the commodities.
Why this strategy worksparticularly well in these environments.
Do you focus on a subset ofthe commodities?
Because you just mentionedweather dependencies.
(18:20):
For instance, gold would notbe weather dependent, but you know,
coffee and cocoa are.
Or are you trading across theboard all the commodity markets or
everything that is liquidenough for your definition?
You kind of answered thequestion yourself.
We only do trade the liquid stuff.
So we are in the four majorasset classes or groups you can call
(18:43):
like metals, agricultural,softs and energies.
And we focus on where theliquidity is the highest good because
that also supplements ourstrategy quite well.
We don't trade some of thesmaller commodities, even though
I know that some of the otherCTA funds you as well, I think participates
(19:05):
in the orange juice market for instance.
That has been a quite a goodtrade for a lot of the CTA funds.
But we do tend to focus on theliquid stuff where we can also grow
our fund quite, quite a bit.
So yeah, so we are the majorcommodities basically, right?
I mean clearly you're tradingway more short term than, than we
(19:28):
do with two to five days, Ithink you've mentioned.
I mean the orange juiceposition for us has been going on
for way more than two years.
So you know, we don't reallyneed to tap into the liquidity of
that market on a daily basis.
But with your tradingfrequency you're absolutely much
more sensitive to that.
Then coming back to the supplydemand imbalances, I mean it is represented
(19:57):
by price at the end of the day.
What is your definition of a,you know, imbalance and how do you
spot that?
Like what triggers this?
Basically we use only pricepatterns as our entry.
We do lean towards veryclassical CTA mythology because I'm
(20:19):
a very strong believer of a.
Of the CTA fund and the corearound the CTA funds modern day trading.
We do need to modify the veryclassical the way that the CTA evolved
and started.
So while we use like a veryopportunistic entry signals, very
(20:41):
classical CTA style breakoutspredominantly we only use the entry
as, as, as the first part ofthe trade.
The real advanced stuffactually comes in after we have the
trade on the books.
And that is our risk management.
(21:01):
One of our core fundamentalsis that we do want to deliver a very
low volatility access to thecommodity markets.
So why we do love commodities,Commodities can be volatile and,
and through the use of AI wecan sort out a lot of the volatility
and create a more linearyreturn profile.
(21:25):
Right.
So breakout to the upside orto the downside that would trigger
the position.
You've mentioned Philip, acouple of minutes ago, weather dependency.
So you're not using weatherforecasts as an input in your trading
signals.
It is really price based.
And price pattern based Ithink is what I understood.
So search sequence of priceevents materializing or the price
(21:45):
making a new high.
And like a trend followingtrader you would then get long once
that, you know, high has been made.
That is true in regards to theweather pattern.
It really comes into the price.
Let's say we have a hurricanethreatening the oil rigs, gas rigs
in the, in the Gulf of Mexicoor Gulf of America, what we call
it these days.
(22:06):
That is actually what we arelooking at because this threat from
the hurricane will trigger aprice movement that we can benefit
from.
And after this initial riskpremium being put into the market,
the market will look at thehurricane development, the speed
(22:27):
the, the classification andwhen it goes landfall and at what
strengths it goes landfall at.
And this period when weactually in this limbo, it gives
us a lot of underlying signalsthat we can use in our machine learning
regard to the momentum,candlestick recognition patterns
and so forth that we use toclassify and use our past history
(22:51):
on the TRA and the trace wehave already made to see if we can
predict the upcoming movementof these particular commodities.
So quite often we doparticipate in these movements where
the market stabilizes a bitand trade sideways.
And this is often enough forus to exit the position capturing
(23:13):
the alpha and then look for anew signal into the same trend if
it continues up.
For instance, is this a fullysystematic process?
Like do your systems pick upthat there is a hurricane potentially
materializing say in Septemberor October and it automatically filters
through to your machinelearning environment?
(23:34):
Or is, is it a discretionaryinput where you need, where you need
to nut your system and tellthe system hey, there's a hurricane
potentially materializing.
You should have a look at this data.
No, it's all price recognition patterns.
So we only use prices input.
There's no manual input inregards to selecting signals.
(23:54):
It all be handled by ourproprietary trading platform.
So we don't use any, anyweather patterns or any manual or
writing in the signalselection process.
Got it, Got it.
And just coming back to theentry signals, you're not forcing
your system to be long andshort at all times.
(24:15):
You can be flat as well, right?
If there's nothing going on,then you don't have a position in
whatever market doesn'tinterest the system.
Exactly, yeah, spot on.
So basically as we only wantto participate in in real alpha periods,
we can be flat and we can beon a partly long allocation or partly
(24:36):
short allocation or combination.
So we don't have any mandatoryallocation towards individual markets
or sectors or groups or directional.
We can be in anywhere themarket is.
Basically let's then move astep forward.
Now you have a position, it'sbeen triggered, let's say a long
position has been triggered bya breakout to the upside and the
(24:58):
expectation is now that you'rein that trade for two to five days.
Is the logic a trend followinglogic in the sense that you would
hold onto the position untilan exit is elected or hit by price,
which is on a reversal you'renever getting out on the top.
Or you have more differentexit techniques where you're taking
(25:21):
profits.
Basically you can divide ourmythology into two groups where we
have the, the systematicapproach, where we do very classical
cta, we enter into a trade, weput in our stops, we use trailing
stops as well.
But this is only one part ofthe trading strategy.
(25:44):
As such, because our majorityof our trades actually is not held
onto the trading stops, we usethe machine learning exits which
takes out, I would say 80, 75,80% of our trades is handled via
the machine learning exits.
(26:05):
So basically what happens isthat on a daily basis we Collect
a lot of data concerning theindividual trade we have in our books.
This could be candlestickpattern recognition, momentum, angle
on momentum and otherindicators as like that is part of
explaining the currentmovement on or the current short
(26:29):
term trend.
We then use this data patternthat we have collected on the individual
trade on the day we go intoour massive massive database where
we can see if we can find asimilar pattern as we have collected
on the day for this particular trade.
(26:50):
And this is where the matchinghappens basically because when we
take this new data put intoour database, it finds a match that
is as close to something ithas seen before.
And as this is historictrades, we do know the outcome of
these particular trades.
T +1, T +2.
So this actually gives us likean indicator on when we have seen
(27:15):
these patterns before whathappened to this particular trig.
And this is basically what we use.
And then we classify them askeep or eliminate.
And if it's eliminate then wetake them off the books and look
for new entries.
Basically.
Are you streaming tick data toyour systems or is it batched into
(27:35):
15 or 30 minute bars?
Or I assume that you're usingmore frequent data than a daily open,
high, low close for what it isthat you do.
We use various time frames,but when we do it on the close in
regards to the machinelearning, we actually put emphasis
into the daily bars and thedaily data.
(27:56):
Because as I mentioned before,we do believe in the CTA mythology,
We do like trends and all that.
So we just classify the trendsdifferently than a traditional CTA.
So while classical CTAs are intrends for the long term, for the
long time, in the long run wetry to capture bits and pieces of
(28:18):
these existing trends, tryingto vet out any counter moves along
the way.
Doing that allows us to maybenot benefit from the entire trend,
but we do capture part of thetrend and we vet out the volatility
allowing us to create a muchsmoother ride than than the longer
(28:41):
term trend followers.
And all the entries entrysignals are pure trend following
signals.
Or do you also combine meanreversion and other type of methods
and strategies into the mix?
We use trend followingstrategies to enter.
The markets trend only.
Okay, so.
So let's have a little bit ofa chat about machine learning and
(29:02):
we're not using it.
I'm definitely not the bestperson to speak about it because
I don't have much experiencewith it.
But how does it work?
Like how do you prevent themachine to learn from the noise which
is a prevalent factor in the markets?
I mean there's also a lot ofsilly things that a machine can learn.
You want it to learn the good things.
(29:23):
I sometimes joke that a clevermachine will probably learn to buy
the breakout and become atrend following trader long term.
But I may be wrong.
So how do you make sure youunderstand what's happening in the
black box?
Or is there even a need foryou to understand?
Or could you just say well wedon't care, we just want to make
money, it's fine as well.
That's a very good question actually.
(29:45):
And keep in mind that we donot use machine learning techniques
for the entry and thatactually answers your concern because
when you use machine learningfor the entry signals then you actually
put great emphasis intoguessing where the market is is moving.
(30:06):
And I'm not a strong believerin in that part actually.
So I do believe that trendshave strength.
So running along a trend is inmy opinion a very strong part of
this environment that we're in.
We only use machine learningas a risk of mythology and this actually
(30:31):
changed the picture quite abit because if you use machine learning
as a risk of mythology,basically let's say if we go back
to the hurricane example weused before.
Market has risen quite rapidlydue to a hurricane threat.
After the first initial riskpremium being put into the market,
the market stabilizes.
(30:53):
Everybody's watching theweather reports the path of the hurricane
and all that stuff.
So the market stabilizes andtrades sideways.
Maybe even some of the traderstake profit on the initial move.
So it might maybe even liketrace back a bit.
As a short term trader oursystem reads the short term trends
(31:14):
towards the long trend.
So these movements and theactions around this stabilizing period
is enough for us to get validdata into getting signals on how
to take when to take off this position.
So we log in the first part ofthis risk premium, take the profits
(31:37):
and then look for a new signal.
So when the market eithercollapses because the hurricane didn't
hit or hit the the oil rigsand cause lot of dis destruction
and the trend continues, wecan actually re enter into the same
trend that we captured thefirst alpha move on and get the second
(31:58):
leg of this particular trade.
So we actually it only uses itas a risk off element which changes
the picture quite a bitcompared to trying to build a fully
AI model that tries to predictthe price targets.
Yeah.
So it's all happening post entry.
(32:20):
Once you have a position on inthe materials Philip that you kindly
shared with me ahead of thecall they mentioned adaptive execution
is the adaptive executionpiece what you mean with getting
out of the position andessentially taking it off off using
machine learning techniques?
Or does it also mean that youhave a more smart, clever way of
(32:40):
executing your trades in themarket using alos given that you
are short term?
You know, I reckon that youhave some sensitivity toward that.
Sure, both.
So we, we both use advancedALOS to get into the market.
But the adaptive stuff is alsovery, very much in meant as the exit
(33:01):
period as well.
We also use like a time of theexits as a parameter as well where
we try to get our exits inliquid periods predominantly.
So of course our stops areexecuted as it happens, but our machine
learning exits actually timedtowards the most liquid time in the
(33:24):
US because remember that thesemachine learning classifications
are not firm predictions thatthe market will collapse or go against
your current position.
It's just a flag saying beaware there's a potential correction
ahead of you.
(33:44):
And this is why we take morein the liquid periods.
Right.
Creating a segue to OLA here.
Ola, I'm sure looks at all thecommodity forward curves, whether
they're in contango orequitation and seasonal and all that.
Does this information triggeranything at Kalkulu or are you maybe
even trading spreads which youknow, you can do in the commodity
(34:06):
markets?
We're not showing spreads as is.
No.
Okay, but Ola, you have allthis information about the forward
curves.
You have this view into thecommodity markets.
But I also know that you're aninvestor in Kalkulo.
What has triggered your choiceto you know, put your money into
a quantitative long shortcommodity strategy as opposed to
(34:29):
one that is fundamentallydriven where you know, we just spoke
about copper and cocoa andcoffee and all the views that you
can have on these markets.
Well, I, I have it as, as, aspart of a general exposure because
I, I like commodities.
I, I've been involved withcommodities for, for more than 20
years.
And, and it's, it's reallywhere I, I, I feel that as I have
(34:51):
an edge but obviously I knownthis the Kakul for many years and
I have been involved for anumber of years.
Probably what attracted meinto an investment originally was
simply the low volatility and returns.
The correlation betweenreturns and the low volatility really
(35:12):
quite appealed to me.
And then there's also simplyto have a different approach instead
of just a long only.
And, and, and when you lookat, when you look back and we look
at some of the performances,it's actually interesting to see
that the, the Bloombergcommodity index that I do that I
do follow obviously on a dailybasis that the performances difference
is not great.
But the road travel to get towhere we are today is massively different.
(35:36):
And that's why I prefer thiskind of strategy because just looking
at the Bloomberg commoditysince in the last five years we've
had three we had several drawdowns.
We had the two biggest one wasa 77 months drawdown between 15 and
20 where it fell 38% then wehad a 33 month drawdown up until
now we still haven't reachedthe top in 2022 but the initial drop
(35:59):
off in this latest correctionwhich we're now trying to reverse
was 22% and that is reallythese kind of drawdowns is, is what
you have to have to accept inorder to get this long term return
which then ultimately end upbeing not being significantly different
to what what Kula has has, haspresented with a, with a much lower
(36:23):
volatility.
So that's really what whatwhat attracted me into it.
And I think also also what Ihave noticed is on, on some of the
months where we've had the, wehaven't had many but I think we're
probably going to get more inthe future where we had to have bad
months for equities we quiteoften find that there's actually
some positive months coming inthrough the commodities.
(36:45):
So also just generallylowering the overall volatility across
your investments.
I mean commodities don't havethe equity drift or anything like
that.
I really think there aremarkets that you should be trading
long and short and when youlook at the BCOM index, well first
off all of these indices, thefirst generation of these indices
(37:07):
they're equity and metalsheavy and then there's all the smart
variations that are morediversified and they distribute liquidity
across the curve more.
But I think COCO for instancebecause it's a relatively small market
is not a BCom member probablyfor the last two years.
They wish they had it in thereit would have made a difference.
But luckily you can say thisyear last year it was the natural
(37:29):
gas which is member that drawdragged it down and cocoa which was
was not involved.
And this year it's, it'sthere's now coffee that is both coffee
and natural gas.
That's that's putting butputting higher.
But yeah no doubt that some ofthe others last year had a very good
year but also meant that therewas a massive re rebalancing in in
the, when the, when theJanuary came because the the percentage
(37:51):
share of or exposure to cocoaout of the total was uh.
Had gone quite extreme becauseof the, the big gains we saw last
year.
So, so yeah, there are, thereare, there are different ways.
I also watch ETFs where theytry to be smart on the role.
It hasn't been.
I don't find them having beenparticularly successful.
They haven't reallyoutperformed the main total return
(38:15):
in the indices.
So even though we are in asituation that you mentioned contanguant
backwardation where wegenerally are starting to see a positive
role yield coming into the markets.
Except.
And the biggest option isobviously where you just have had
your profitable short innatural gas.
Because.
Because the contango structurebasically means this is a market
that even though we're seeingsometimes seeing big rises over time,
(38:38):
you're actually making money.
You have been making moneybeing short simply on the roll from
a low to a high contract every month.
Well that's part of it.
We're not rolling every month.
We're aware of seasonalities.
But yes, for sure we didbenefit from the generally upward
sloped forward curve there.
(38:59):
Philip, I think you wanted tojump in and say something.
Yeah, that's right.
Just to supplement Ola'scomments here.
One of the things that weactually also have seen in the Calculo
fund is that we over the pastseven years almost since inception
has actually delivered equitylike returns at two thirds of the
(39:21):
volatility of the broad stock indices.
This is also just one of theselling points for Kalkulo strategy
being a low volatilitystrategy and non correlated, non
cyclic.
So actually allows a lot ofour investors and we see that a lot
when we, when we meet ourclients that we can actually offer
something different to themthat they can get from other asset
(39:44):
classes.
We are liquid, we're notcorrelated and stuff.
So it's really an add on fromfor many of our clients in this space.
Sure.
One question I'd have PhilipBenjola is you're domiciled in Denmark
and I think you're the onlyfund that I know that is domiciled
in Denmark.
How did that come about?
I mean it is not known as ahedge fund land.
(40:08):
Neither is Germany for sure.
But you know people go toCayman or Luxembourg Island.
You're in Denmark.
Why that?
First of all we are Danish.
Sure, that is one thing.
But yeah you're definitely right.
I think we might have broaderaudience in a different country than
Denmark but sometimes havingthe space on your own is also an
(40:31):
advantage.
So we are probably the onlypure commodity fund in Denmark and
it's our job to educate and toget the word out there.
And anything that's on thehorizon for you guys in terms of
how you're looking to growyour business and develop it, maybe
become more international.
(40:53):
Definitely.
We have explored the optionsof doing like a US setup because
strategies like the Kulostrategy would really thrive in the
US I believe.
So this definitely one of ourthings on the agenda to look abroad
to see if we can actuallyexpand our client network in more
(41:19):
commodity knowledgeable areasas well.
Would you also trade SMAs orare you purely focused on managing
funds?
We'd be looking into both, actually.
We wouldn't rule out SMAs thatwould require a US license for that,
of course, but definitely it'sall in the pipeline.
(41:42):
Excellent guys.
Thanks very much for comingonto the podcast.
I really enjoyed it.
I hope our listeners will findthe conversation valuable and interesting
as well.
As usual, we'll include themost important points of today's
discussion in our show notes.
And also, as usual, youshouldn't hesitate to contact us
if you have any questions.
Our email is infooptradersunplugged.com so thanks again for
(42:06):
listening.
And until next time on TopTraders Unplugged.
Thanks for listening to TopTraders Unplugged.
If you feel you learnedsomething of value from today's episode,
the best way to stay updatedis to go on over to itunes and subscribe
to the show so that you'll besure to get all the new episodes
as they're released.
We have some amazing guestslined up for you, and to ensure our
(42:27):
show continues to grow, pleaseleave us an honest rating and review
in itunes.
It only takes a minute andit's the best way to show us you
love the podcast.
We'll see you next time on TopTraders Unplugged.
Popular Podcasts
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Therapy Gecko
An unlicensed lizard psychologist travels the universe talking to strangers about absolutely nothing. TO CALL THE GECKO: follow me on https://www.twitch.tv/lyleforever to get a notification for when I am taking calls. I am usually live Mondays, Wednesdays, and Fridays but lately a lot of other times too. I am a gecko.
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.