RAOUL PAL: Keith, good to get you back.
It has been a while since you have been here. I have been on set at Hedgeye, but you
have not been on Real Vision for a while, so always good to pick your brains. I want
to do something a bit different today. We will come on to markets and your views in
a bit, but you spend a lot of time talking to people about investment processes and what you
guys do at Hedgeye. I would love to you from top down to getting to a more granular level in
terms of when it gets to position and stuff, how you go through that process. I think we
will just help people just to understand. KEITH McCULLOUGH: I think that this is what at
least I feel differentiates what I am trying to say and how I am trying to educate every single
day, I really do not have any predisposed thought on where economic gravity should go, or what
country your asset allocation I should be long or short on the equity or the fixed income
side, or what asset class from commodities, to anything, really, I can even talk about
like investing as in wine as an asset class, which I think surprises some people. The reality
is that if there is a time series, and there is a long standing time series of that and I can back
test whatever that asset class is, then I am going to have a view on it based on incoming data.
What we do at a very basic level is that we measure and map the rates of change of
growth and inflation globally. We do that for about 50 countries, we keep that live, we have
about 175 ETFs, that we are also measuring and mapping in what we call signal terms. You have
two-track process there, you have the measuring mapping of the economic data, and you have the
measuring and mapping of the signal if I just call it that. The signal generally is frontrunning
economic quad, or the economic data. The quad, of course, is a two by two model. Its growth and
inflation. There is four different permutations in rate of change space that those
can be quad one, two, three, and four, with Quad 4 being the dangerous one, where both
growth and inflation slow at the same time. That is very damning for equities, for example, in
those countries, and it is very good or generally good for the sovereign bond in that country.
There are certain asset allocations you make, which we would call conditionally based on the
economic data. Every single day, I get new data so like a good Bayesian boy, or what we call Bayesian
inference process, which are, it really should not be a surprise anybody. It is really like bean
counting. Every day, I reserved the [?] right until somebody in this country, socialist or
not, takes it away where I can change my mind. What really changes my mind is back on
the other track, which is the signal, and the signal is a price volume volatility
signal, which really weighs on the relationship of those three factors. A lot of people
look at the surface area of signal, or what I affectionately call moving monkeys. If you
are just looking at the moving average of a price, you are looking at a one factor model
on the surface area of say the water. Now, if you are looking at the current, the
rocks and take it a little deeper into what is actually happening to that body of water to use
the metaphor, price, volume and volatility, then you start to get a little closer to where I am at.
What I really find the signal speaks to a phase transition back to the quad, a phase transition
in natural space or in thermodynamics, everyone would know that mean, we go from one thing to
the other thing. What really signals is the vol of vol or the volatility of volatility, when that
starts to change, asset classes start to change. That is really just frontrunning what we
call a quad. Those are the two big parts.
RAOUL PAL: Are you using implied vols when you are
looking at the vol of vol, or are you looking at realized volatility and over what time period?
KEITH McCULLOUGH: Well, they look at realized across durations. That is another important
thing about the signal. I think a lot of people-- again, everyone does it differently, but I
do it consistently. I have done it my own way for a long, long time. When I look at it,
I look at it across durations. NICHOLAS CORREA: Sorry for interrupting your video, but I
have a very important message to share. At Real Vision, we pride ourselves on providing
the very best in-depth, expert analysis available to help you understand the complex world of
finance, business, and the global economy. So if you like what you see on the Real Vision YouTube
channel, that's just the tip of the iceberg. You should come to realvision.com and see how we're
not leaving any stone unturned. From publishing more in-depth videos, live discussions, written
reports, and our latest feature, The Exchange, where you get a chance to engage with experts,
fellow subscribers, and learn from everyone's experience, which can't be wrapped in a video.
It's an experience which you live and learn from. So if you go to the link in the description or go
to realvision.com, it costs you just $1 to get a month's access to this incredible content. I don't
think it's something you can afford to be without. I am looking multifactor. Multiple things in
macro, everything in macro all at the same time. I am looking at across durations. I
call them trades, trends and tails, and I am looking at realized.
If I look at the trade duration, for example, which is three weeks or less, I could have one
day, 10-day, two weeks all the way up to three weeks in that model. I am looking at realized vol
relative to implied vol. Whether or not implied vol is trading towards a discount versus what has
been realized in the last 30 days, for example, or what's trading at a premium. That is really
important, Raoul, because you can see that is where the market is betting on future volatility
or not. There is a lot in that information when paired with a probability weighting of the
next economic quad, or the economic reality. I know that is a lot not to you, but there is
a lot to that. What we are really trying to do is simplify the complex there, as opposed to me
just saying, hey, what is this bloke over at prime broker J, B or C think about market positioning?
Those days are over, like getting those types of tidbits or talking points. Now, we can
measure map all of it, all of it. You could not do this before, whether it is on
economic data or on the signaling of it all and so we had modern computing power. I
think that that is one of the big reasons why we have become more accurate, really,
since 2016, is that I have more information faster and I can see where the crowd is tilting
the wrong way at precisely the wrong time more clearly, and that is an important point about
looking at realized versus implied in particular. RAOUL PAL: Let us talk a little
bit about that. When you have got realized trading below implied over whatever
time period, let us say it is a trading signal we are looking at. I do not know what
time period we are looking at, but let us say whether it is two week or one month
or seven-day, whatever you are looking at, what is the signal you are looking
for? What gets you to sit and go, that is something interesting. Then we will talk
about how we use volume and the other stock with it. That signal is the vol of vol, what is the
kind of signals you look for that make you think, okay, something may be happening here?
KEITH McCULLOUGH: Well, it is a very short-term tactical thing. Some people try to reduce me to
just being a trader, which is fine, I take it as a compliment these days. What I am really trying to
do is identify a particular point in time within the full investing cycle, the full investing
cycle is the sine curve. With an economic growth, it is peak the cycle on growth, or inflation, or
vice versa, sit in the bottom. That makes me out there in the longest term, but then I zoom in
to the very immediate term and one good example on implied vol, the way to use it really, is
there is a two-factor model there as well. When a market is trading at the low
end of its risk range, for example, what typically happens is you get a big implied
volatility premium or big spike in implied, people are buying protection after something goes to the
low end of the range. That is a pretty good spot if your quad setup says to buy that thing, to buy
more of it in size. Conversely, if something is, A, at the top end of its range, so I publish
these daily, it is at the top end of its range, and the implied volatility discount digs deeper.
That is a great example of complacency and/or capitulation. In other words, the bulls are fat
and happy, they are very happy because the price is up at the top end of the range, and bears have
covered their shorts and they have capitulated. That would be a good spot to take off some of
the position or short something of that position. RAOUL PAL: Talk to me about risk ranges,
because I see you put this out a lot on Twitter, your risk ranges. How do you define
risk ranges? Is it some volatility style structure? How do you look at risk ranges?
KEITH McCULLOUGH: That is built within the same prism and all that, you can have a risk
range on any duration, but the daily risk range is the one that really matters, because that
is the-- think of it like the strike zone. Now, some people who maybe never made it to
the major leagues think that the strike zone is always the same. Well, it is not because
there is a person called the umpire behind you, who has an opinion on that. Again, until they
fully automated, the strike zone can move, or is very stale like one of my partners here says.
Like football, you can draw something up in the training room, but the trash cans actually
move in real life. That is the risk range, it is moving proportionally to volatility
changing. What people have a really hard time observing is just that very basic point. If
volatility rises, on average, what happens is your risk range widens. Your standard deviation
of outcomes starts to become tougher to call, you got a drunk umpire. Think about it that way.
Conversely, when you have a tightening risk range, what's happening is that the volatility on both
in intermediate term basis and on a trending intermediate term, three months more basis, is
falling and makes that asset or the asset class one that the machine or asset allocators
want to flow to, because that is what you learn if you study economic history
within the lens of market returns, is it flow like on an intermediate to long term
basis goes towards things that have falling, low regimes of volatility that are falling,
and they love rising regimes of volatility, high and rising. That is, you got to use
it all together and I think that that is where-- sometimes people are following it in terms
of how I see it, it is very clear in my own head, my challenge is actually to communicate it to
other people that is why I spend so much time on. RAOUL PAL: The risk range is essentially the
volatility band is like, I do not know what, one standard deviation band over that time period
or something that gives you the probability that there is a short term trend exhaustion, or
if you are using a longer-term indicator, it may be a longer term exhaustion, is that what
risk range is to you? As you said, within the volatility range you are in, it has got to the
top end to where it should be for that move. KEITH McCULLOUGH: Yeah, and it
is a nonlinear equation. Again, it embraces uncertainty and embraces moves that
move the move if you are looking at Treasury bond volatility, for example, the MOVE index. Again,
it is never one standard deviation. In fact, it is quite the opposite. What happens is that a move
in vol of vol, in other words, Benoit Mandelbrot taught this, which is, if you study volatility
historically, you will find that it is one, episodic and non-trending. Again, volatility
moves are typically to be faded, in particular, if you are in a bullish trend of a market.
What happened, what really matters is when that episodic thought to be non-trending cluster
of volatility, which blows up your risk range, appears, well, whoa, that might be the leading
indicator for a phase transition economically. That is what I am really looking for, like when
you really get into the screws of the rescaled range analysis in Benoit Mandelbrot's beautiful
book, it is second to my Catholic Bible in terms of my bookshelf or rank order, and I definitely
will have to put God ahead of that. Again, Mandelbrot would say that I am
standing on the shoulders of gravity, and God would probably agree with that, too.
When you look at, if you read his book very closely, I have read that book so many different
times, and retaught myself through every single mistake I have ever made, I really built it
using, if you want a hint on how to build it, that is exactly how I built it. Again,
what happens is that you get a cluster of volatility that becomes just like anything
in fractal math becomes the new trend. Now, those are the big calls, Raoul. That means
that we are going to a new economic quad, that means that it was not episodic and non-trending,
it is something you should be acutely aware of in that particular moment of cycle time.
RAOUL PAL: We will come into the quad stuff in a minute. Okay, you have now got the vol of vol
signal, let us start at the trading moment first, and then we will go into the big picture. Trading
moment, you have got the vol of vol shift, you then said there is a couple of other factors
comes to play, volumes and something else? What? KEITH McCULLOUGH: The price, volume and
volatility. It is the interplay between the three. That is the basic three-factor model.
RAOUL PAL: You see the vol of vol shift change. You are looking for the price change. Then
what? You are looking for a pushing volume, or how do you put [?] with that?
KEITH McCULLOUGH: That is an important question. It is the relationship much like I was using the
analogy of a riverbed, or the water and the rocks and the speed of the water underneath. Again, when
price is going up, what you will often find, and you can find it uniquely within
each signal, when the price is going up and the volume is going up at the same time,
that reflects there is a higher, higher volume of water, higher pace, higher conviction, more
people believe in it with no change in volatility, it is typically in a low volatility environment.
When you get a move down in a bull market in price, and the volume is decelerating and
still, the volatility has not changed, it is a buy the damn dip signal. If you
get a down move in price on rising volume, on a new change, a potential cluster of volatility
that will be, could be the beginning of a new trend, that is the one to pay attention to on the
risk management side. You see what I am saying? RAOUL PAL: Yeah. Now, we have a rough idea
of how you look at the trading signal. How do you then bring that into a longer-term
signal? I guess you just do longer term variations of it, longer-term volatility, but
then how does that fit into the quads? How would you get all of these different disparate price
signals, and say, actually, we have got a shift in our quads? Firstly, talk a little bit about the
quads, and then this transition between one and the other and how the asset prices tell you that.
KEITH McCULLOUGH: Again, two feet on the floor, God willing in the morning, and this is hard, you
have to pay attention. This is why I love being-- you can call me a trader full cycle investor. I
do not care. It takes a deliberate study. You have to pay very close attention not to just that one--
you have to do that through everything that is big in the model that matters. You have a three-factor
model, price, volume volatility, let us say between 25 and 30 factors the machine considers
as causal at any given time, those things can be things like copper, the 10-year bond yield, the
10-year bond yield versus the 10-year bund yield. RAOUL PAL: This is all Bayesian?
Things shift around all the time. KEITH McCULLOUGH: Yeah, always changing. Again,
deliberately studying it, writing it down after each model reloads for me in the morning, I get
my signal, then I say, what is the aggregate of that signal saying about the secret to the
universe, which is the quads? The quads, again, is calculus. The quads is the rate of change of GDP
growth real and the rate of change of inflation. Without a debate on specifically what government
is lying to about either of those in terms of how they are stated, it is explicitly focused on
the rates of change of both of those things. Again, secret to the universe to my friend Steven
Strogatz, Professor Steven Strogatz, who wrote a great book that teaches people that are not
mathematically inclined, but they are interested in things like gravity, and that the earth is not
the center of the universe, like some people think valuation is or a 50-day moving average. Again,
we are trying to move the discussion to the place that you are standing on the shoulders of giants
of science. Again, when these things start to move all at the same time, Raoul, they start to quite
frequently signal that we are going into a phase transition into a different economic quad, with
the most dangerous one being deflation, Quad 4. For example, what would happen to the dollar,
or when the dollars start going up, the 10-year yield starts to go down, that is a similar set,
for example, in fractal math as we would say, but a similar set that is quite causal in Quad
4, if it is non-episodic, it is trending, it is an up dollar and down bond yields, I am blowing
one of the things in the big thing down, but that is one of the biggest similar sets of that.
RAOUL PAL: Let me get back a bit, when you have this whole bunch of signals that you say,
you download the model in the morning, how do you know whether there is enough signals to say,
hey, there is something changing in how we should allocate within the quads now. Because obviously,
the economic data does not change very often. How do you know when the markets are giving you
a real signal that there is enough of a signal, you would say, I can do something about this
versus, that is interesting, I will just watch it. KEITH McCULLOUGH: Most of the time, it is like
what the model would call a Brownian motion. It is nothingness. That is the whole point, and
Mandelbrot taught us this too. It is about the particular things, it is not the average of
things and it is not about the feeling of things, it is certainly not about the valuation of
things, valuations is not a catalyst. Again, when the probabilities all line up together,
you act, and it takes some time to wait and watch on that. What is easier to do is once
you are in a new trend is to just ride it. For example, from middle of 2018 until May of this
year, the dollar was signaling global deflation, alongside global bond yields, just to use
that similar set that I just mentioned. I had a lot of conviction, because
every bloody day that I wake up, Raoul, it would do the same thing, it would say the same
thing for two and a half years. That is easy, the hard part is to realize when that was saying
that in the middle of 2018, or in May of 2020, I guess it was more like two years. The hard part is
recognizing, and that is where I have my struggle. My daily struggle is not against somebody's
talking points, or what somebody wrote in the FT, it is against the machine. It is against me,
myself, and I, my flaws as a human being, my susceptibility to being sucked into to a fake
signal, or a head fake, or episodic non-trending signal. There are a lot of those. That can
drive you a little crazy, so you have to tone yourself, you got to wake up sober in
the morning, you just start with that. RAOUL PAL: How are you dealing with economic data
nowadays? Because economic data is a bit odd right now. How are you putting that into your model?
Considering things are moving very fast, everyone, we are all trying to use real time economic data
as well, how are you adapting to the new world in terms of the economic data when you will build it
out of the quads themselves and the big signals? KEITH McCULLOUGH: It is an important question,
too. We are constantly back testing whether it be through what we thought were former and
interesting AI algos, like random forests, or whatever we are using. In the background, we
are always trying to find like, what could be a better data set? There are a lot of data sets. We,
as a research team, we have 40 different analysts, we have 20 fully loaded with data scientists
and software engineers, 20 other guys and gals doing that, so I am looking at a lot of shit.
What you find is that a lot of shit means nothing. What I found in the last, since 2016, even the
Chinese data, the Chinese data, which is got to be the most made up of data, apples to apples,
they have the same people making up the same shit. What you find is that the rate of change of the
made-up numbers are quite causal. We got long China on what we would call a quad one, quad one
is when you have the rate of change of inflation falling, as real growth is accelerating. Being
long China has been in a beautiful position, irrespective of your political opinion of the
Chinese or your position on China long short. Again, the data itself-- or Germany, for example,
which is just, and I know you want to get into thoughts on positions and what not later, but
Germany just started to signal just a raging Quad 4 in Q4 of 2020, which made German equities a
bigger short than anything in the US equity space, in particular, and made German bunds a long at
like 40 basis points on the 10-year German bund yield, which would drive the valuation
expert bananas. Again, I actually think the government data is been-- if you are
doing it, and it is not about the data that everybody has wrong. It is about what you see.
I am not trying to be Picasso. I am obviously Mucker not Picasso, but you do have to get in
there, and you got to see is the data confirming or denying your position? Is it rotten data?
It is infrequent that we are seeing that, if anything, I have started to buy countries
in the last three months like Finland, their equity market that I did not know that we could
nail Finland, but the model nailed it. I think thanks to the Finns for their government data.
RAOUL PAL: Now, you got the cards. We have talked about that. You got your economic data,
you got your positions, how do you know which position to take and what not to? How do
you filter what is an opportunity versus-- you can spit out a lot of trade, but you need
to know what the conviction is going to be on a trade or whether you should take the trade or
not. How do you go through that filtering process? KEITH McCULLOUGH: I had a wonderful discussion
with Diego Parrilla at the Hedgeye investing summit, that you did such a great job at, and
everybody appreciated the time you spent with us and people, that is what they really care
about. Thanks for spending some time on my process because that is really what I am spending
with Diego and all of a sudden, he is talking my language. He says, you know what, Keith, you
know what a lot of people make a mistake with when it comes to running a portfolio
away from understanding that the goalie and the defenders matter in winning the
game, and it is not just about having scores to use maybe a Bitcoin analogy, people like
the idea of having nine Maradonas on the field. The thing is going to triple and quintuple,
that is great, we have to have that. What we really need to do, like the way I think
about it, my own net worth and preserving and protecting that so that I can compound returns is
having a full team of players, a fully allocated, full investing cycle team on the field, and then
grossing up the bets where there is an asymmetry in the payout. Asymmetry and the payout. Then
Parrilla said this beautifully, because most of Wall Street, they have been taught to think of the
asymmetry in conviction, like what is this guy's narrative? What is this guy's story? When I listen
to him or her, do I feel like they feel? There is a lot of feel there. I feel absolutely nothing.
When I looked at TIPS yesterday, the most asymmetric thing that I could have done yesterday,
which is the most boring thing to a lot of people, is allocate assets to TIPS. Sell things then
booking gains in certain things like Bitcoin, which is going up, and reallocate to TIPS.
It is about pruning and planting. That is another way to think about it. When you can
go anywhere in your strategy, I actually think saving and making money gets easier. Where I
think this game gets a lot harder is when you are betting and forced to bet on Diego Maradona
having nine of them on the field, and they will have to score three goals. That is not how I do
it. Never have, never will. I know it is only for a certain type, but I do think it is for a
lot of people who would prefer to not have 30% to 50% to 80% drawdowns of their net worth.
RAOUL PAL: Yeah, and that is why I am very interested in what you do. It is more about when
the model comes up with a lot of positions, what do you do? Just say, everybody, well, here is like
20 new positions, or you say, okay, we filter it down to 10 where we think there is more asymmetry,
or we see the vol of vol change at a different rate, or the volumes confirming it better? How
do you filter the trades down to something that you said, listen, guys, these are the ones?
KEITH McCULLOUGH: I have predetermined rules on what my max allocation is by asset class.
12% would be a currency position is my max, which would include gold, because I considered
it a currency. 10% would be a fixed income max position. If I were to take a 10% position, or
position that I want to go big in, in TIPS, for example, the max would be 10% of that. Equities
would be a 6% position. Commodities, which I had put Bitcoin in that bucket or any other asset,
which includes wine, I would make that 4%. When you add all that up, you actually do not
need that many positions, when you go big on, we are talking about a dozen positions are really by
asset allocation. There are a lot of other things that could become my bigger asset allocation.
That is the other thing about, as you know, anybody who has run money knows, to make money,
you have to have something on to understand it, feel it, trade it, risk manage it from. In my
speak, I need to watch that sucker in price, volume, volatility space, and then that is what
gives me the go button. Guess what, newsflash, every time I hit the go button, it does not work,
but if you probability weight your decisions, and you are right 65%, 70%, 75% of the
time, it is a great way to manage your own money and/or other people's money.
RAOUL PAL: When you are looking at that overall allocation, so you have got
this maximum risk allocation, right? KEITH McCULLOUGH: Yeah, by position.
RAOUL PAL: What are you normally running when you are looking-- you are observing the recommended
portfolio over time, what kind of risk allocation do you normally have? Do you say, listen, we are
running at a 30% risk allocation across the board, or do you say, hey, listen, we
are running a 30% allocation in bonds, or how do you think about the whole?
KEITH McCULLOUGH: I think about that through the lens of the quads as well. We have written
extensively on this, some people are like, well, I need white papers. I say, well, it is on
a piece of white paper, I am not going back to college to go prove that out. Again,
what you find is in Quad 4, for example, you have a 60 plus percent allocation to dollars
and fixed income. Whereas in Quad 2, you are going to have the opposite of that, you may not have
any fixed income, or I will be short fixed income. Outright, I will be short gold in Quad 2 as well.
I think about it is conditionally affected obviously, by the quad. That is why,
again, it is a go anywhere strategy where, in Quad 2 for example, in 2018, I was short
gold. Then I went long gold in Q4 of 2018 when we got to Quad 4 signal, and the Quad 4
signals started with dollars and treasuries, just thinking about the aggregate of the signal.
That is what I did. It will be catchy, it will look totally different than some random 60/40
bullshit model is what I will just call that, because that is just not a sensible way to do
it. Nor is targeting net exposure, or the only thing that I would use is target the max gross
exposure by asset allocation, and that is also a function by the way as a function of volatility.
The reason why I say that Bitcoin is a commodity, or it is an asset and commodity and asset are
the same thing, is that Bitcoin's volatility, historical volatility, resembles, if you
have, I am sure you have looked at it, it looks very much like a commodity, or in
some cases, an equity. That is why I put it in the non 12% bucket as a currency, I do not think
it is a currency anyway. If it had currency attributes from a volatility perspective, I
would not care, I just put it there. It is a very dispassionate process in terms of scoring
the realized volatility of an asset over time. RAOUL PAL: When you are dealing, let us say,
with the Brazilian real versus the Swiss franc, two wildly different volatility currencies. Do
you volatility weight them, or you do not, or you would equate them? How do you think about that?
KEITH McCULLOUGH: That is a great a great question and an important point from a risk management
perspective. I should have said it at the outset, each position is beta adjusted or volatility
adjusted. If I take a short position in the real, to use your example, against
the dollar, that is going to be a smaller notional position than shorting euros
against the dollar. Because euros got less volatility against the dollar than reals do.
RAOUL PAL: Good. Okay, so I think we got a decent idea. I think there is a lot people-- people have
to take a few notes and learn some of this stuff. Now, we have gone through that, let us talk
about where we are now. Even I am a bit confused, because there was a-- you were infirming
Quad 3, which is the-- explain Quad 3. KEITH McCULLOUGH: Quad 3 is where you have
economic stagflation. That is where you have a down dollar, you generally have a central
plan to inflate asset prices in that, in this case, the devalued dollar. Quad 3 versus
Quad 4, stagflation versus Quad 4 is deflation. RAOUL PAL: From what I can see, you are firming
Quad 3 for a long period of time, now this whole rally up did all of that, then there has been a
bit flipping around here between Quad 3 and Quad 4, what is going on right now? Yeah, how does
that work when it is flipping around for you? KEITH McCULLOUGH: Well, what is slipping around
are the signals in the data, and this is where people who want me to be an elegant mucker
are wildly disappointed, because there is no typical holding period or amount of
time and space, as Einstein would explain where your inability to move starts to matter.
That is generally through the lens of volatility. Again, what volatility does is it shortens time
and space, when it is rising, when volatility starts to fall, it allows everybody to be
a long-term investor. That is what it is. Quad 3 to Quad 4, we are talking about-- and we
are very precise in terms of how this is measured, you basically went from a probability of it being
Quad 3, 4, basically, let's say since May until August, that was a very what we would say a high
probability, 60%, 70%, 80%, to moving towards, oh, Quad 4 just became a 30% probability. Then in the
middle of September, it actually got up to 57% probability of Quad 4, surpassing Quad
3. There was a point in time there where I went back to long dollars, long treasuries
and shorting the NASDAQ, which is what you do, for example, in Quad 4, and then boom this past
week, it flipped right back to now. It is like it is a pretty close race, by the way, it is 49%
for Quad 3 and 47% or thereabouts for Quad 4. RAOUL PAL: Here is an interesting question
for you. When you get a big transition shift, do you often find this volatility between the
quads, the vol of vol of your own signals picks up? I am guessing there is some informational
value in the fact that you are flipping around, that there is often a change of regime.
KEITH McCULLOUGH: Oh, yeah, there are big time. Typically, the government wants to infuse their
thoughts on it, and that is indeed perpetuating the volatility. Think of something simple
like a stimulus bill, or a not stimulus bill or a stimulus bill that does not happen on the
timeline that you are positioned for therefore you got to position for the next timeline.
Again, you will see that in implied vol space, and that is where 30-day implied versus realized
starts to flip around, because it is not my model that is flipping around. It is people that
are flipping around. It is people, it is human beings that are buying and selling options.
RAOUL PAL: The market is clearly scrambling to understand, okay, what is the narrative happening
right now, what does this all mean? Which is, as you say, the market does not make its mind up
and so you are seeing the model change a lot.
KEITH McCULLOUGH: Right, and some people are super
uncomfortable with me doing that, and I love it, because embracing uncertainty, like, I am happy
to be a mucker and a trader because you have to be a little less intelligent, a little more numb to
having the answer to all of it all the time. You have to be able to say, I do not fucking know,
and you know what? The market does not either, that is why I do not know. When the market
knows on my metrics, price, volume, volatility, and knows what quad it is, I can make money on
the long side. Like, it is so easy. It is so easy. When it gets hard, there are a lot of things that
the market is trying to discount on different durations. It is not just hard for you. It is
hard for the whole market. I like periods like that because I love trading vol in particular, and
I have a process for it. Imagine the alternative, Raoul. You are sitting there without all this
information that I have, and you are hostage to your feelings, I would get crushed in that
environment. Like if all I did was trade on valuation, my feelings and what political
party I thought should win, or what stimulus should work or not. These are pretty dangerous
thoughts, at least for me. That would kill me, I would lose so much money trying to go with that.
RAOUL PAL: Yeah, I know. I get it. We got this volatility in this transition shift. It is not
clear, makes sense with what is going on. The market is one way or the other, bond yields go to
the top of the range, they go to the bottom range, go back to the top of the range, they are clearly
struggling to figure out inflation, deflation. That makes sense. What are the strongest signals
you are seeing? Amongst a lot of this noise in the quads, you are obviously seeing some strong
signals. We talked about the Ibex earlier in Spain that you have been short for a while because
you have got a consistent Quad 4 signal. Talk us through some of those consistent signals you got.
KEITH McCULLOUGH: The most consistent signal for going on three years in the case of
shorting the Ibex or Spanish equities is A, the duration of the quad in Germany and Spain,
and the market signal that supports them. In other words, that is why German bund yields went to like
minus 66 basis points last week, that is a super strong signal. Even CNBC started talking about
it. That was the day to sell some bunds. Again, because bunds get to the top end of the range, or
the yields at the low end of the range, that is a very easy one. We have been short,
anything European equity on the southern side of Europe has been fantastic.
What has also manifested as quite a strong signal, like I said, is the north of Europe like
we are long Norway and Finland, again, this is not a theme. It is just what the signal--
that is very typical of what I do. I will tell you a story about what is happening after it is
happening. I do not tell something to happen and therefore, I am. That is not how we do
it. That is a big signal on the European side. In the US, I think the most contrarian signal that
continues to manifest is a real potential cycle, a full investing cycle bottom in commodities.
I think that that is a big one, and that is one that I continue to broaden my bets in. Obviously,
Bitcoin has been one of my bets and I have traded around it, do not tell anyone. Again, I trade
around cattle, I trade around a lot of different things, corn is making a new high, all of ag
looks fantastic to the upside. That is despite the dollar having this maybe I will go back
to deflation, but really, what it is saying is that by the time we get to 2021, our cleanest
projection is that we are going to be USA Quad 2. The difference between Quad 2 and Quad
3 is Quad 2 has more inflation faster, and you actually have real growth. That would
not be hard to envision because next year in Q2, you are comparing against a pandemic. In
other words, if you are looking at inflation, no shit, you are going to have inflation, the
comparison is a negative oil price last year. The question really, I think the market is struggling
with in terms of having a clean direction is, A, how high is inflation going to go if you have
MMT in the US, and they actually do commit to burning the dollar from a very high level.
As you know, the dollar is at really, it is at a full quad cycle high. That is what it is.
That one, to me, looks like the one that I am most likely going to press and stay with, is go
back to shorting dollars and buying commodities or shorting-- China looks fantastic to on the
long side or EM is starting to broaden, which is the same thing. To get the dollar right, you get a
lot of things right, including EM and commodities. RAOUL PAL: They are all the same trade. Dollar
weakens and EM goes up. It is as simple as that. As you pointed out earlier, generally the
most hated EM goes up the most, which is China. I was looking at the MSCI China chart
and I am like, fuck that looks amazing. I am like, I have no idea what the narrative
is. The dollar looks really good right now. KEITH McCULLOUGH: Yeah, the China consumers CHIQ
is the ETF that we prefer there. We also like KWEB, which is China internet, KBA, which
is EA shares. That is the more boring of the three but they all work, and they are working
for the right reasons. You have had a lot of domestic consumption occur and I do think that the
country's moving, this is more thematic. Again, I tell myself the story after it happens, is
that countries are looking inward. Therefore, domestic growth, consumption growth in
China has actually been quite strong. Retail sales just went from zero to plus 3.3%.
Industrial production growth, sub seven. Yeah. What if they are made up? You know what
the US number is on industrial production? Down seven. It is also relatively-- I like it
when a growth story is relative and absolute. RAOUL PAL: Talk to me about-- because a lot
of ETFs that you look at in your universe, some of these are deeply illiquid. How
do you deal with liquidity within this? Because you have got different
clients, you have got institutional, some of them retail, how do you think
about liquidity? You do not look at it. KEITH McCULLOUGH: I know that-- what has been a
humbling experience here is that, for example, like the Teucrium corn ETF, the ticker is CORN,
or SOYB, same manufacturer of ETFs, the CEO pings me because I hit a real timer buy signal on
it for the first time and the volume quintupled or whatever. He is like, well, who are you? Can we
talk? Like, you do not want to talk to me, there is going to be a day that I short your ETF, buddy.
At the end of the day, that is humbling, because as we call them, Hedgeye Nation acts in concert
when they see something big, like they have not seen me go bullish on commodities forever.
Something like that, it was not like you do not, in the very short term, like maybe in the first
one or two hours, you might have a bit of a disconnect versus the underlying, but the
price of that ETF and the price of corn, they are not mirrored, but they are pretty
close enough for me. Let us put it that way. RAOUL PAL: Going back to these quads now-- again, the wrong question to ask you,
but what do you think is going to happen looking at the data you have got? Are we going
to stay in Quad 3? Are we going to Quad 4? Are we in the hands in the air and fuck it, let us
see what happens phase? What is your best guess? Again, I know you do not like guessing, but I am
just-- you know the nuances of your own models, you know the feel of how that usually plays out?
KEITH McCULLOUGH: Yeah, it is not unlike watching the behavior of your children. Yes, we would like
to huddle our children for the long term, but in the very immediate term, we can have some
big behavioral problems with these children, and over the intermediate term, we are tasked
with smoothing them out and going with the highest probability decisions that we can
make to help them get to where we would love to huddle them for the long term.
Again, I am into this whole huddling lingo because there are certain things I do love,
and they certainly are not assets that I trade. If you think about it that way do not forget that
I have-- maybe I should have said this at the outset as well. All of our subscribers for a long
time now have been getting what we call Quad Map. The quad map shows you 50 countries and what
the next four probable quarters or four quads by quarter, economic quadrant is what a quad means
by quarter. That is the story I always go wit, Raoul. That is the story I tell myself. When
I look at China today, or Taiwan, which is an even more interesting example, the next four quads
for Taiwan are Quad 2, 2, 2, 2. I know what that means. That is buy AWT or buy Taiwan.
When I look ahead at the US model, I think we are going through the last of
the slowing, if you will, because this is a recession on a year-over-year basis. The
debate is between deflationary recession or stagflating one. Stagflating one is much better
for gold, for example, than a deflating one. The probability is high and rising
that we are going to be in Quad 2, and maybe we are going to be there for three to
six months. I do not know. That is the part that I do not know. Again, God willing, every day,
I am going to look at the signals, anything in the data that changes the quads.
RAOUL PAL: When would you think that transition would make the most probable to
occur into Quad 2? Is that Q1 or Q2, or earlier? KEITH McCULLOUGH: Could be. Well, in terms of the
way that a practitioner would price it or me when I buy it, when would I go from buying? I went from
being long Quad 4 basically for a six-week period. Then day to day, I was even flipping around
which again, drives people crazy, but I love it. You could have it, you could envision a scenario,
let's say Trump wins, and the market does whatever it does and goes to the level that freaks people
out or not, that might be the moment to put on the Quad 2 portfolio in November. It could
be sometime in December when people are tax loss selling or tax booking, taking their tax
gains because they do not want to pay x rates. I do not know what that date is going to be. That
is why typically when I set up for the next quad, Raoul, I always say this, I do not know what
the fuck I am going to do like until like, literally t-minus one to three days.
RAOUL PAL: It may not happen, but [?] as well. KEITH McCULLOUGH: Yeah, I know what I am
looking for. I know what I am looking for, but it is just like hunting or fishing, you do not
just get to pick your prize. You have to wait on it. This whole concept that you can just like roll
into it and stay with it, I just do not do it that way because I am a selfish bastard, I want to get
that price. The market setup right now is really interesting from a US equity perspective,
because there is so much bubbliness to it all. You have a very high regime of volatility,
like NASDAQ volatility is at 35 today, which is basically untenable if it were to stay there.
A bull has never made money with equity vol staying with the three in front of it, or in the
30s. Let us just be clear on that. It could also mean that the shape of the VIX curve or the shape
of the VXN NASDAQ volatility curve is basically got all the fear you are going to ever want into
it ahead of the election. There is two different ways to look at everything, and I just wait and
watch. I just say, okay, play yourself out. I am not going to tell you what to do, but when you
do it, I am going to, going back to my children. If you do that, I am going to discipline you. If
you do that, then I am going to compliment you. RAOUL PAL: Looking across this slightly
confusing world that you have described to us, what are your strongest convictions? Clearly,
short Europe seems to be still pretty strong. Sounds like long developed Asia
seems pretty strong in equity terms. KEITH McCULLOUGH: Yeah, I like those because
I am long short too, do not forget. Having them against each other has been great.
I think that the hardest one to solve for is going to be the US equity market, and how to
time the components. Because if it is Quad 2, you know what you buy? You buy my shorts, you
buy financials, and you buy energy stocks. That is the one when you think about the asymmetry
of payouts, the worst two sectors in Quad 4 are the energy stocks and the financial stocks.
They have been Quad 4, and even buying oil for that matter. Those are the things that if I am
going to go big on something new, those are the two that I could have a lot of conviction in,
but I have more conviction in waiting right now. Riding commodities broadly, as I said, is
an easier one. Because we already have a lot of things, which I do think are linked
probably to some, whatever you want to call it, Green Deal and stimulus like trillions at a time.
The electrification of even the hair that you and I have left, we are going to electrify everything
so we are going to need natural gas, copper, I believe in that theme, because the market
signal does and that is one that I am in fairly significantly in terms of exposure right now.
RAOUL PAL: How do you deal with a market that is as bifurcated as this? You got banks, oil, a bunch
of stuff that is screaming Quad 4 in your terms, and then you have got a bunch of stuff that
is not at all, like commodities. When you have got the market that has got a complete
psychosis about it, how do you make a decision, or do you just play them as like,
okay, fine. You got split personality, and we just tried to split personality.
KEITH McCULLOUGH: That is pretty much it. I do not ask questions in the morning when I buy
more natural gas when it is at the low end of the range, and why do not I buy oil when technically
I should be or from a quad perspective, I should be long both in t-minus a month or two. I just
do not care. That is why I really do not care. RAOUL PAL: I am thinking of the signals that he
gives for the quads itself. Your macro framework is when you have got the split stuff, it
is fine when it is the Ibex and the DAX are all falling, and bunds are falling. It
is pretty fucking obvious with that one. KEITH McCULLOUGH: It is easy.
RAOUL PAL: That market is doing two different things at the same time, and it is screaming two
narratives. Those two narratives are inputs into your model to give you quads, is that one of the
reasons it keeps flipping around with them, maybe? KEITH McCULLOUGH: Well, yeah, well, because
we were flipping between Quad 3 and Quad 4 and financials are shorts in both. That is a good
example of one that you would be short financials in both Quad 3 and Quad 4, you would be long
utilities in both Quad 3 in Quad 4. There are certain sector styles that work in both quads.
I think that that message has been very clear. No matter what you think it is, Mr. Market, if it
is Quad 3 or Quad 4, which are two very different things, we are going to buy utilities,
we are going to buy REITs, and we are going to short financials. That has been clear.
The big move that I am talking about is when it is neither of those two quads, and it is Quad 2, that
is where you short utilities, short REITs, short treasuries, and buy financials and buy energy
stocks. That is the one that is confusing in terms of how to time it, but not because of the quads.
We are not just going to go to Quad 2 because a certain long only manager needs to rotate. They
are getting rotated this year on that idea. RAOUL PAL: Look, it is fascinating, and we will
see how it plays out to see where we end up. I am really interested in the self, in the volatility
of the volatility of your quads. I have a feeling that some informational value within that, where
the market seems to be fighting with itself to decide, okay, what is the narrative that
it wants to trade? I do not know how long this volatility of this plays out for, my guess is
it is election related. We will see after that. What happens if we do not get results
or what happens if we get split Senate? Nobody is thinking that anymore. Everyone
is like, okay, it is probably a Democrat victory. There is a bunch of other people say, well, I am
sure Trump will nail it. Nobody has really ever-- that narrative, obviously, we do not get anything
for three months. That would be fun to the market. KEITH McCULLOUGH: Yeah. There is so many
asymmetric risks. Again, for me, I do not feel-- like I really sleep fine. It is like when those
payouts appear, the asymmetry of the payout appears, I am comfortable because I have a map not
only in terms of what-- and I think that that is a key thing here too, Raoul, is that some people
say, well, the guy is cocky or confident or both, I am just aware. I do not think you can wake up in
the morning, I would not play this game that way, being unaware of what literally what economic
quadrant, any country you are trading, or any commodity exposure you have, any bond you are
trading, sovereign bond being the primary ones that have been talking about in this
conversation, but you should not pay me a thing to say anything about anything until I
know everything looking backwards, because the best leading indicator for the future is what
happened and rate of change terms in the past. That is what gives me conviction to act when
I act because I have a map that is based on the time series of what already happened.
RAOUL PAL: Brilliant. Keith, thank you so much for your time. I think it is just really good
and interesting to dig into how you see things. Because you spend half the time on Twitter, trying
to explain why you do things to people, because they do not understand. I just think it is good
for people to understand how you do things because you did something very differently. I think it is
unique approach and it is a systematic approach, which I think is useful and very appealing to a
lot of people. Keep up the good work with what you are doing and let us see how this all plays out as
ever. We can all talk, as we said, we can all talk about the future. The future is not anything until
it has happened. We will find out what happens. KEITH McCULLOUGH: Oh, thank you. Thank you.
I appreciate that. Not everybody asks all the questions you did. I think that is the first
part of understanding somebody's processes, trying to understand so thanks. I sincerely
appreciate you taking the time to do that. RAOUL PAL: Alright, Keith, take
care. I will speak to you soon. NICK CORREA: I hope you enjoyed this special
episode of the Interview, the premier business and finance series in the world. However,
this is just the tip of the iceberg. For more in-depth content and expert analysis, visit
the membership link in the description, and unlock a week’s access for only one
dollar. This dollar could change your life.
