Andrew Lapthorne is the Head of Quantitative Equity Research at SocGen, a role he’s held for nearly 14 years.

Given the breadth of topics covered by bank research, it should be no surprise that this conversation takes some wide swings as well. We discuss everything from thematic baskets to style premia and machine learning to ESG.

One of my favorite parts of the conversation is when Andrew discusses his research into strong balance sheet names in U.S. small-cap equities. For all the depth in discussion of how index composition rules affect small caps, why Merton’s distance-to-default correlates to credit cycles, and how this trade can potentially be a positive carry hedge, I love that the inception for the idea came from just updating spread sheets.

While this podcast goes wider than it goes deep, Andrew’s experience allows him to sprinkle a bit of wisdom in every topic we hit.

I hope you enjoy this episode with Andrew Lapthorne.


Corey Hoffstein  00:00

All right 321 Let’s jam. Hello and welcome everyone. I’m Corey Hoffstein. And this is flirting with models the podcast that pulls back the curtain to discover the human factor behind the quantitative strategy.

Narrator  00:20

Corey Hoffstein Is the co founder and chief investment officer of new found research due to industry regulations he will not discuss any of new found researches funds on this podcast all opinions expressed by podcast participants are solely their own opinion and do not reflect the opinion of new found research. This podcast is for informational purposes only and should not be relied upon as a basis for investment decisions. Clients of newfound research may maintain positions and securities discussed in this podcast for more information is it think

Corey Hoffstein  00:51

This season is sponsored by simplify ETFs simplify seeks to help you modernize your portfolio with its innovative set of options based strategies. Full disclosure prior to simplify sponsoring the season, we had incorporated some of simplifies ETFs into our ETF model mandates here at New Found if you’re interested in reading a brief case study about why and how. Visit with models and stick around after the episode for an ongoing conversation about markets and convexity with the convexity Maven himself simplifies own Harley Bassman. Andrew lap Thorn is the head of quantitative equity research at sock Jen, a role he’s held for nearly 14 years. Given the breadth of topics covered by bank research, it should be no surprise that this conversation takes some wide swings as well. We discussed everything from thematic baskets to style premia, and machine learning to ESG. One of my favorite parts of the conversation is when Andrew discusses his research into strong balance sheet names in US small cap equities for all the depth and discussion of how index composition rules affect small caps, why Mertens distance the default correlates to credit cycles and how this trade can potentially be a positive carry hedge. I love that the inception for the idea came from just updating spreadsheets. While this podcast goes wider than it goes deep. Andrews experience allows him to sprinkle a bit of wisdom in every topic we hit. I hope you enjoy this conversation with Andrew lap Thorne. Andrew lapthorne Welcome to the program. This is I think you’re gonna be a really fun episode for listeners, because we’re going to talk about a whole breadth of things. I mean, the fun thing about talking to someone who’s a head of Research at a bank is you see a whole lot of stuff and have seen a whole lot of stuff. So again, really excited to record this episode with you. I want to start off though with a little bit of a fun piece of trivia about you for the listeners, which is that one of your sort of first forays into the world of quant and algorithms and computer science was actually in designing and optimizing a transportation system for a set of ski resorts. And so I’m very curious, thinking back to that experience, what was some of the most challenging parts of the problem? And how did you solve it?

Andrew Lapthorne  03:14

Well, it makes it sound a lot more sophisticated than it actually was. I left university with a computing degree in the early 1990s. And there was a recession on so I thought I’d go skiing. I thought my prowess at skiing would interest the tourist ski company I was going to work for but they were mainly interested in the fact I spoke French and knew how to do computing. And actually turns out my skiing wasn’t that good, either. So they basically had a problem, which is they had 27 resorts, they had free airports. And they were shifting quite a lot of people around with a fairly limited amount of buses every weekend. And so it was my job to try and work out how to get all those people into all those buses into their resorts reasonably speedily. And I’d like to say it was a sophisticated algorithm, but it tended to be quite a large sheet of paper, which was start off with all the resorts and all the airports. And then what were the problems? Well, Snow was a major problem, because it would snow and you’d have to rip it up and start again, and fork because Shaumbra airport over time had a bad habit of closing. So it was incredibly challenging, better work, because you had this whole idea of how the world was going to pan out. And then things literally changed from the start of the day throughout. I mean, I used to get a call. I think the worst call I ever got was when essentially all three airports I had all closed, so we had to send the airplanes back. And the difference I guess is that when you pick a stock it doesn’t get angry because you picked it. And obviously we’ve tourists they do get angry if things don’t work out quite so well.

Corey Hoffstein  04:54

I feel like all the stocks I pick tend to seem to get angry at me but maybe that’s just my experience. Uh, let’s fast forward a bit. So you now sit in the role of Head of Research at sock Jen. And one of the things that I think about very often is the world of investing. There’s just a huge breadth of different topics that can be tackled. And so I’m curious how you think about structuring a research agenda for your team over time, how do you pick what the team should be focused in on

Andrew Lapthorne  05:25

could probably ask my team, whether I structure things at all, actually, I mean, first and foremost, I work with some very talented, enthusiastic individuals who are looking at all different parts of the market, they’re looking cross asset, they’re looking long term, they’re looking at the market microstructure. As you save as so many topics, which we’re trying to look at at any one point in time. There’s topics we would like to look at, particularly when it comes to new data, new techniques, machine learning, etc. But at the same time, we’ve got a very wide and varied client base, who has its own problems. And we tend to, we obviously drive our own agenda, what we want to talk about, but a lot of what we’re talking about is stemming from what our clients are concerned about, and problems which they’re facing. And then you’ve got the market dynamic, which always comes and shocks you. One minute, retail investors are massively important. Another minute people are worrying about liquidity because of a particular hedge fund has got into difficulty. So yes, as you say, there’s always something to write about and focus on. And that list never seems to get any shorter would seem. So yeah, so we have a long term agenda, which gets upset quite a lot by the short term.

Corey Hoffstein  06:43

I’m curious as to how bank research has evolved over the span of your career thinking back over the last 1015 20 years, how has the type of research that you’ve done, or the areas you focused on meaningfully changed?

Andrew Lapthorne  06:59

It’s changed a lot. I mean, it’s gone from, from my perspective, I mean, when I started, we didn’t even have email for a start. So we used to actually just write research go to the printers get printed, and people used to just walk it around and manually to our clients. And I remember actually updating models and sticking them on floppy disks and sticking in an envelope so they could get walked around to clients. Apart from the technological side of it. In terms of topics, he very much evolved from this kind of fundamental stock picking or fundamental investing perspective where you were providing information and ideas into a fundamental process to becoming increasingly systematic, not just in the case of designing systematic strategies, but thinking about the difficulties systematic investors might be having. So a lot of it becomes more trading focused, you know, the cost of putting things on, we have quite a large focus on ETF research. And we have a business, which is solely dedicated to analyzing benchmarks, ie, what’s going to happen to a particular formatic benchmark, what’s going to happen to the MSCI benchmark that has become a kind of niche product 15 years ago to one of huge focus. So in summary, I think just the impact of non fundamental investors has really changed, I suppose the range of things we talk about and the range of things we write about,

Corey Hoffstein  08:41

I spend a little bit of time on that idea of the growth of index investing, both truly passive index, and then maybe sort of these smart beta rules based or thematic basket type index products that are out there. And you’ve seen a dramatic growth over the last 20 years and accelerating growth really over the last 10. You mentioned that it sort of shines an increasing light on the importance of the index research team. I know recently, I was reading some of the research from that team, and they were highlighting some dramatic index changes happening in ESG. benchmarks, I was hoping maybe you could spend a little time talking about what that team is looking into, and the types of insights they’re trying to uncover in their research.

Andrew Lapthorne  09:25

I think one of the big difficulties of going from, say like a purely passive benchmark, where the benchmark is largely constructed around concepts of size and price, so your traditional market weighted benchmark, which already if you look at a typical benchmark, there’s about 5000 index changes a year in a passive benchmark, which kind of really negates the whole concept of it being passive. The amount of things that you have to look at if you’re managing a passive benchmark is quite extreme, but that’s been around and whether For a long, long time, but what we’ve seen over the last I suppose 10 years is lots of indices based around other concepts. We’ve seen smart betta turn up. And we’re increasingly seeing for Matic benchmarks turn up. And that leads to other questions. Because if you’re investing purely on the basis of a theme, when do you stop investing? At what point? Do you say no, if you’ve got a market cap weighted benchmark, everything just grows. And it tends to be quite a broad benchmark. But if you’ve got a something based on an idea, at what point does that idea become too big of watch time, does the flows into that idea become problematic. And we know the response, I mean, the index providers then have to respond by changing the structure of that benchmark. And most, I suppose, providers of those type of products will always say it’s a nice problem to have i You’ve exceeded your expectations. So we’re looking increasingly at that we obviously produce as well, quite a lot of thematic ideas. But there’s complexities in that, because often you’re trying to squeeze too much money into a particular idea at a particular time. And that’s obviously very much in the headlines with respect to some of the particular ETFs. At the moment,

Corey Hoffstein  11:16

I want to stick with this notion of thematic baskets for a second, I’ve noticed that a lot of bank research does tend to circle around this idea of publishing these different thematic baskets that might be relevant over times. So for example, last year, there were a lot of banks publishing different sorts of work from home baskets, or reopening baskets, or inflation sensitive equity baskets, from a quant research perspective, how do you sort of see the role of these thematic baskets versus, say more traditional factors and styles?

Andrew Lapthorne  11:52

I mean, as a bank, we produce thematic ideas like other banks and other providers, from a quantum perspective, it’s harder to do, because we’re data LED. So what defines quantitatively what defines a work from home basket? You literally sit there all peloton, okay, great, we’re gonna go. So that’s the trouble, you can’t back test the idea at some point in time, you’re saying this is a viz of a particular stocks which suit this particular idea at this particular time. So it’s very difficult for the systematic investor to embrace that type of process. Now, you’ve got lots of new datasets which are turning up, obviously, ESG is the big one, you’ve got low carbon as well. And we’re doing our best to try to work with all this complicated new data, which is often contradictory to produce baskets, which aligned to that theme. But even there, the data is changing over time, the methodologies are changing, the regulation is very much evolving. So it’s a real challenge, I would say to the systematic investor to embrace fanatics, I think where we can come in is where you’ve got a formatic. And you want to overlay something sensible on that for Matic. So maybe you want to add a set of valuation criteria to it, maybe you want to add a set of risk characteristics to make sure that you’re not getting too much drawdown in the crisis. So I think at the moment for Matic is operating somewhat independently of factors. But we’ve actually had quite good experiences combining new datasets with factors. And other people do the same, you’ll call it the pure factor approach or whatever. Just because you’re investing in a theme doesn’t mean that you should neglect all the other good investment criteria that you might have. So we have pure fact that ESG benchmarks which are optimized towards ESG, that the ESG comes with a dataset which has biases, which are not helpful. So they may be regional and sectoral biases. And maybe they’re too expensive. Maybe we got negative profitability, all that type of thing. And what we found is that once you realign that dataset, or that formatic, with your equity factors, actually, it works far better. I mean, what you find is actually the new data set doesn’t add much value, but that’s what the clients want. And it’s your factors, which are doing the heavy lifting, and helping performance. So I would say, on a 20 year view, we’ve always produced for Matic baskets, every sell side is always come along with these great ideas. I think the difference now is a greater demand for them. It’s coming from the client base, both institutions and retail and maybe that’s because we need to be more interesting. If you say I’ve got a 6040 mix. My bonds are yielding nothing. And I’ve got some s&p, how do you differentiate yourself? How do you come up something which is new and exciting whether something new and exciting is going to make people money is a mute point. But I think a lot of the product innovation is coming from the lack of choice in standard assets. To be honest, I’m staying

Corey Hoffstein  15:07

with the thematic theme for a moment. You and the team have been writing a lot about inflation lately. And I think inflation is a really interesting one from a quantitative perspective, because for the most part, it’s been a non factor for the last 20 years. Most quants working today really haven’t had to deal with significant inflation, at least in most developed markets. And so it’s been a non issue and a place where there really isn’t a lot of data to work with. Given that there seems to be a growing concern about inflation going forward. How are you in the team approaching trying to tackle this risk? And maybe you can talk about some of the things that you’ve learned?

Andrew Lapthorne  15:46

I agree with you on the 20 V view. I mean, I think I’ve been doing this for, you know, since the mid 90s. And I think I’ve had my first serious question about inflation risk in the middle of last year. And that was related to supply chain disruption, because of COVID. I’ve also had the honor of sitting next to Albert Edwards for pretty much all of that time, and Albert’s Ice Age has been very much a deflationists deflation view of a world with yields heading to zero. And he’s been absolutely correct on his fixed income call and that dynamic. And I think that’s what we’ve all grown up with, for the last 25 years ever lower yields deflationary forces. But we’re at a juncture now, where fiscal policy is increasingly using the magic money tree or modern monetary theory, depending how you want to call it. And we’re at a point where the market has been driven by valuation change, it’s been driven by lower bond yields. It is very rare to see a stock market move for so long, solely on valuation change. I mean, the amazing numbers for me is that if you look at the progression of MSCI World profits over the last 10 years, they’re the same, they haven’t gone anywhere in 10 years. Now, of course, we’ve got these fantastically great stocks out there, which have done very well and very much in the news. But there’s a whole load of stocks which have not done so well. Now, we’ve ended up after a long period of quantitive, easing and central bank interaction with the markets where defensive assets are expensive, you know, secure assets are hard to come by at a reasonable price. Now, if you then think about this in another way, if the risk free rate is 5%, how much risk can you take? I would suggest that you don’t mind so much losing 20%. If you could chip in the risk free rate at 5%. If your risk free rate is naught, how much risk can you take? And I think this is a modern misconception between kind of market practitioners and economists or maybe central bankers. Central bankers say, Well, if we decrease a risk free rate, everyone will want to take risk. But we don’t really want to take risk. We want to be in a risk free rate. We want five 6% Vol. So we go find around looking for instruments, which have five to 6% Vol, we’re not ever going to buy anything with loads of vol. So during the era of QE, anything with low vol or anything apparently safe as rerated massively, so you’ve got a whole range of negative yields, you’ve got quality stocks, which were very, very expensive growth stocks, which are very, very expensive, and anything which is volatile and dangerous. cyclicals maybe financials dirt cheap. So the market is perfectly positioned for bad news. And then good news comes along. And we’re like, oh, let’s hope that good news doesn’t lead to higher bond yields. Let’s hope it doesn’t lead to the Fed putting up interest rate charges. So I find it are we talking about inflation risk? Are we talking about the interaction of inflation risk, forcing a change in interest rate policy earlier than people expect? Or maybe just an interest rate policy, which is entirely incompatible with quality stocks trading on 25 times. And just as we could rewrite equity, because we’re pushing people out of bonds while we could direct equity as bonds become more attractive. So the inflation story for me is massive, not because I necessarily believe that we’re going to get sustained 345 percent inflation simply that any inflation above what we have is incompatible with essentially the state of any expensive markets. I think one of my preferred charts is just looking at the real yield on a nominal sorry, yield yield on a global portfolio. So if you just take global equity, global fixed income, so global sovereign bonds, and a bit of cash and a bit of credit, and you minus out inflation, you’ve got a negative real yield, and that’s before you start charging for running that fund. Now, obviously, you can make capital gains but that on the assumption that you then sell assets to make those capital gains. So what is your real return from holding assets permanently? Well, it’s negative at very low inflation. Obviously, if inflation goes up, it goes more negative, but actually what really will happen is you’ll start losing money on your fixed income instruments. So it will it go even more negative. And so I think that’s the context. You know, and I think fat is a reason why everybody is so concerned about it. Inflation is not high, we’ve still got output gaps, we’ve still got unemployment, all economists see, and I do agree with them, there is no reason to believe this kind of disinflationary scenario is going to end necessarily anytime soon. But even if we get two years of sensible nominal GDP growth, the assets that you’ve owned over the last five years are probably incompatible with that.

Corey Hoffstein  20:49

You guys have published a couple of notes on that front one I actually think I just read this morning before this call, which was talking about the role of cyclicals versus rising yields GDP growth and how there really haven’t been periods of continuous rising yields over the last 20 or 30 years to give that sort of tailwind to the cyclicals that they’ve just remained meaningfully undervalued. I thought what was very interesting, and one of the prior pieces you wrote on the inflation trade, though, was recognizing that inflation sensitive equities, if you had bought those as a basket have largely just gone sideways for 15 or 20 years. And so the question becomes, if you want to buy this basket, and even try to quantitatively analyze which securities you should be buying, it is potentially a dangerous trade from a relative return perspective, if that inflation doesn’t materialize, but I thought you guys put together a very interesting trade where you almost tried to synthetically create this call option. Using that basket, I was hoping you could maybe talk about that a little.

Andrew Lapthorne  21:50

Yeah, it’s a very difficult sell. Really, if you sit there and go, Hey, here’s my back test, it’s delivered naught percent return over 20 years, and it suffers from 75% drawdown. It’s quite, it’s quite a tough chart to look at. But of course, we suffered from the opposite. Holding inflation assets has been a bad idea. But we like the idea of having an inflation exposure in a multi asset portfolio because it hedges all these other risks. Hedges is duration risk, essentially. But the nature of the asset you’re buying is probably benefiting from a supply and demand imbalance. So if you think about it, copper prices high, you’re making a lot of money, all that supply at some stage is going to come on stream. So these assets do tend to be quite violent, because you always have this mismatch between supply and demand. And we know that all the supply tends to come on stream at exactly the time demand is going in the opposite direction. So you’re faced with an asset, which has got fantastically useful diversification properties, but something that you really don’t want to hold on a permanent basis. So we thought we’d overlay a kind of short term trend following strategy on this basket. Now, if you think about it, when you sell a call option, you have to hedge it using the underlying and you’ll hedge it based on the Black Scholes formula. And as it goes up, you’ll start buying more and as it starts going down, you’ll start unwinding now. Sandrine and Gauri who I work with wrote a great paper in 2013, basically demonstrating that short term trend following strategies, human factors trend following strategies, pretty much have the same mathematics. As Black Scholes, you’re basically positioning, your position is a function of the trend and the volatility of that trend. Now, there are certain strategies, which benefit massively from gap risk. Now one of the problems with value investing. And the reason why buying and holding call options on value investing is difficult, is because they’re very expensive, because they are volatile. And a lot of the return comes from things like the Pfizer announcement where you make 15% overnight, so the call option is going to factor that in the cost of that call option. So if you were to systematically buy call options on value, yes, you would avoid the downside, but you’re just a road through the premium. So what we wanted to do was to have something which looked like a call option, where you’re picking up some of that upside, but you’re dramatically reducing the downside. Now, luckily, unlike a value strategy, an inflation strategy tends to trend. So as the strategy is moving up, you’re accumulating more, and there’s sufficient short term trend in that strategy for you to make enough money. So I think it’s a really interesting concept. And yeah, I think it’s something that we want to evolve further. I mean, clearly, it’s not something it’s not. It’s not an easy thing to do. I mean, you’re trading the thing every day.

Corey Hoffstein  24:51

I want to jump from sort of thematic ideas to styles and factors in one style that you have been writing about for For years now, is this idea of buying strong balance sheet names, and very much in particular in US small cap equities? How did this idea come to you? And why have you put so much focus on it?

Andrew Lapthorne  25:13

I think a lot of our ideas come from updating spreadsheets. I mean, I still update tons of spreadsheets, because I just think it’s useful. I know. Because you see the data, you see the data evolve, and you get used to what the data should look like. So for example, I’ve been updating us reporting account dates for decades. And normally, debt piles should grow larger. As the business expands, you’re borrowing to finance the business, then, again, putting the blame squarely at quantitive. Easing and central banks is that essentially what you had is you had weak demand. So there was no need to borrow money and build a new factory because you didn’t have the demand to justify the new factory. So all went into share buybacks. There’s lots of people who get very emotional about share buybacks. But I think we all agree that borrowing lots of money to buy back your own shares is leverage. It’s not a return to shareholders, we could all go and borrow money at the bank and stick it in our own bank account and go, all I’ve done well. And so we saw this spectacular rise in debt through the 20 1314 period. Meanwhile, cash flows were fairly static. And that started raising alarm bells. And then we started thinking about it started looking around, and it seemed us small caps, were getting very much involved in this. So you had an asset class us small caps, which have very high levels of leverage. And that is something that we wanted to turn essentially into a proxy trade on credit. We’ve all seen The Big Short, we’ve all seen the amounts he was drawing down waiting for things to happen, you could identify a credit problem fine. A lot of people did identify the credit problem before the financial crisis, putting on that particular trade cost you a lot of carry, and shorting credit costs you carry so like any kind of hedging strategy, you’re you’re bleeding costs waiting for it to come into fruition. Bonds. Handily in US Small caps, you’ve got a real benefit of avoiding junk. AQR have written about it, I think Beco more recently have written about it, we were writing about it the same time, you don’t have the survivorship bias that you have in the large cap indices, where junk is naturally removed through the rebalancing process, and eventually drops in market cap and it ends up in the small cap space. So junk is quite a big drag on performance for small cap indices. And I completely agree with the work which has been written, which is if you clean up the small cap index for that junk, you’ve got a little bit of alpha. So now that allows us to have basically long small cap x junk short, small cap gives us a small alpha. So now we’ve got positive carry, waiting for credit markets to get into trouble. So we’ve created a synthetic credit hedge. And that works really well, as well, you know, balance sheet risk tends to then work against you in a recovery. Because the junk flies, normally, you’ve had intervention, credit markets start improving volatility starts going down. But at the same time, you’re making money on all your risk assets as well. So that shouldn’t represent too much of a surprise, good quality companies will normally underperformed bad quality companies during an inflection point. So yeah, it’s been a it’s been a long story that and I don’t think it’s gone away. The leverage hasn’t gone away. credit markets are very tight. I think people still worry about credit, but it’s something that people will start worrying about, I guess when interest rate rises, come back into view.

Corey Hoffstein  28:56

One of the things that I thought was really interesting about the trade was a quality when you ask someone to define quality, there’s all these different ways that they do it. And I think if you go to that AQR paper, you mentioned that quality versus junk paper, they use a composite of a number of different ways of trying to measure the quality of a business. In the research you’ve focused on. You’ve looked very specifically at Mertens distance to default. Why did you guys hone in on using that characteristic specifically?

Andrew Lapthorne  29:25

I knew nothing about credit. 20 years ago, when I was very lucky enough to a guy joined our team called Sebastian Len Chetty, who worked in the credit risk department. And he introduced me to distance to default. Why did he introduce me to distance to default, because it’s one of the key metrics that we use to measure credit risk. And essentially what you’re saying in this model, which is Robert Mertens model, it’s been around since the early 1970s. It’s essentially saying you’ve got the value of the business, what is the probability of the value of that business being built over debt? that you’ve got in that business if there’s no equity left. So if you think about its price, like an option, you’ve got the current price, you’ve got the strike price, which is a debt, and then you’ve got a whole range of outcomes. And the volatility of an asset is pivotal, rightly or wrongly, we measure the future worth of the business based on its volatility. So, the reason a utility company can borrow more than a semiconductor business is because the assets are far more stable, you would expect a regulated assets to be less volatile than a highly cyclical asset. So the amount you can borrow historically, at least is contingent on essentially the volatility of your business. And I used it particularly in the 2000 to 2003 kind of cycle where I was trying to differentiate between cyclical beta and leverage beta, one of the key inputs to Mertens distance of all is volatility, you could be a cyclical business. So you could be volatile, and you’ll find your beta is fairly consistent through the cycle, you have a beta of two on the up and up to beta of two on the down. If you add leverage to that, you get asymmetric beta. So you get a beta of two on the way up and you get a beater of five on the way down, you also get this vicious cycle element where Mertens distance default and things like Moody’s can V are used to approximate credit risk. So what then happens when your volatility starts going up, your credit spreads, start widening your CDs price balloons out and that freaks everybody out. I always used to say that equity markets and credit markets were basically two people waving at each other in a building, they never speak to each other. But as long as the other guy is happy, they’re happy. You just hope someone doesn’t turn up grumpy because then it’s game over. It’s really important when it starts to think about drawdown risk, the connectivity between credit and equity markets, for me is very much related to volatility. It’s model based. At the moment, you’ve got this big debate where you’ve got credit spreads, which are very, very low, and everybody who’s positive on credit thinks volatility is too high. And then you’ve got other people saying volatility is fine credit spreads as being overly helped by intervention. So there’s quite a gap there. I know you indulge in distance to default, in some of your charts,

Corey Hoffstein  32:30

indulge is a good word for it. In a prior conversation we had were talking about this topic, you mentioned the phrase to me that the circularity of connecting volatility to credit is a really useful framework for thinking about risk. What do you mean by that?

Andrew Lapthorne  32:48

For me, when you start looking at credit risks and credit models, are you spend my time prior bank talking a lot to that I used because we were quite bearish, we talked to the credit risk department. And they talk in that language, they talk in the language of trying to assess what the risks are across our whole platform. And they’re not just talking about single stocks, it’s talking about all kinds of things. Essentially, they’re all using some measure of what the future worth of something is going to be. And therefore, equity volatility, therefore becomes Pivotal, because that’s probably the thing, which is most life, if you’ve got the book value of something, it’s not really going to help you. And that then permeates through the system. So when people say, okay, the credit spread is narrow if the volatility is also very low, but you see that they’ve been piling on the leverage, nonetheless, what’s probably suppressing the credit spread is the volatility of the asset, not the fact it’s got low leverage. And what people then forget is that individual companies volatility is not connected to the individual company, it’s connected to markets. Essentially, what you’re doing is you’re kind of selling your soul to the credit, devil, really, you’re now at the mercy of volatility. So you’re looking in the background, just what’s the leverage on this business, maybe even what the volatility of that business is, but now the future downside risk of that business has materially changed, because for every point, additional on vol, is going to have a very different effect than it had done before. So I think for me, it was just really allowing me to understand just how circular and dynamic credit markets are. Now, I think the authorities know this. I think this is why in a crisis, one of the things that authorities have tried to do over the years is suppress volatility enough to reopen credit markets. And I thought it was fascinating last year because we’ve had a pandemic. So we’ve got something which is not solved by interest rates, although everyone tried to solve it to a certain degree with interest rates. So the authorities had to do Something to improve credit market. ie they couldn’t just suppress vol. So they just directly intervened in credit markets. So I think when vol goes high, ball goes high and a crisis effectively closes down the credit market to reopen the credit market, you then need to suppress vol. Or to put it this way in a crisis, no one knows what anything is worth anymore, it then becomes very difficult to raise money on the back of very volatile assets.

Corey Hoffstein  35:26

So how do you reconcile this idea that the back test of sort of going long, strong balance sheet names, and short the index or even short the junk names within the small cap universe creates this positive carry trade creates this alpha, but it’s also negatively correlated to credit shocks, which are positively correlated to sort of economic shocks, you shouldn’t be able to sort of have a positive carry hedge? How do you sort of reconcile that idea?

Andrew Lapthorne  35:55

Oh, I think it’s a reversal effect. I mean, we’ve seen it very much in the last year, um, you saw it again, coming out of the 2008 2009 crisis. I mean, you do not want to do long, short quality, your short quality leg, as you saw in the Russell 2000. Last year, with retail investors participating in stocks, which were heavily shorted, I mean, that’s unmanageable. So we short the index to make it more manageable. But you are going to get periods like we’ve seen over the last 12 months where the lower quality stocks fly, because what you’re doing is you’re putting all the bad stuff. That’s happened. volatilities improving credit spreads and narrowing, you’re removing the problem, the problem becomes quite intensive. When you’re having a strategy with a, which essentially has a long short quality bias, particularly in the last six or seven years, you’ve ended up with a short value bias. And while that may seem very good, most of time, you don’t want to be short value in a recovery. I don’t think the strategy is sensitive to that. But it’s not wholly sensitive to it. So yeah, no free lunch, I’m afraid.

Corey Hoffstein  37:07

Well, as it relates to value, you said to me, quote, value as a portfolio of problems, but a call option on good news. Curious what you meant by that, and what you think the ultimate implications are for value investors.

Andrew Lapthorne  37:20

I think we now differentiate why people like to quote, Ben Graham, and stuff we’re not when we’re picking up portfolio of value stocks, when our value investors were systematic value investing, we were buying stocks, which looked cheap on a variety of metrics. And we could play around with those metrics, but we ended up all with a fairly similar amount of names. So it’s not like people suddenly surprised that this stock is cheap, and no one’s noticed, we’ve gone through the report and accounts with an old pencil and identified some intrinsic value that no one has seen before, you know, so, essentially, we are buying a stock which has a problem. Now that could be idiosyncratic. So you could have a problem connected to something which has gone particularly wrong for that particular company. And actually, personally, I quite like buying idiosyncratic stocks, maybe to counterbalance the fact that my business is doing it systematically. But when you start building a portfolio of 200 300 names, you’re buying macro problems. As Brexit comes into view, UK stocks, D rates, you’re going to be picking up more UK assets. As the cycle peaks at the beginning of 2018 people are start you’re going to start picking up cyclical assets, you kind of start picking up energy assets, eventually, you start picking up stocks, which have suffered most from the pandemic, you now have a portfolio which is exposed to macro news, which is completely unpredictable. You know, whether that’s a Shanghai G 20, whether that’s a Pfizer announcement, whether it’s I still don’t know what people said on the 12th of March 2009, I’ve got a whole list. And all of a sudden, this stuff flies. So essentially, what you’re doing is you’re putting downward pressure on stock valuations. And I’ve always said, you know, as my cheesy anecdote is that you’ve got a strong spring, you’re pushing problems on that spring, the tighter that spring get, the more acute, you have to make the problem. And you don’t need the problem to go away. You just need for it to get less acute. How do you measure the strength? How compress that spring is? Well, it’s valuation. So we’ve shown that a lot of the returns to the value factor are highly concentrated around these turning points, which we’d all like to time but we can’t because the turning points tend to be some kinds of unpredictable announcement. So essentially, this is why we think it’s a call option on good news. You’re building up a portfolio problems which you think are sufficiently discounting the problems, not that they won’t go lower or because when you buy it, the problems not going to go away. But through the cycle, you’re going to benefit more from buying those problems when they were cheap i the problems that you’re buying, which is typically the economic cycle to be fair, or transitory, I think that makes it quite a good right hand side hedge. The trouble is, when you go to buy a call option on it, the person selling the call option knows that and therefore makes it far too expensive to do. But I think if you’re buying a stock, which is really cheap, and you don’t know why it’s cheap, is a problem. I mean, you want to see the problems in the stocks, because if something’s trading on 10 times, and you don’t know why you’ve probably missed something fairly important. So yeah, so that’s how I see it. Now. The other adage is, why do you then do it systematically and ignore macro? Because you don’t know when it’s going to be? So yeah, you hold a value stock for two years, because it pays you out in five days, you just don’t know when the five days are. There is a sound logic for not doing market timing on value, because it’s unpredictable. You don’t know what the announcements going to be. You just know that you’re going to be basically buying problems which are temporary

Corey Hoffstein  41:13

sock gents quantitative research team started its endeavors into machine learning and stock selection in 2016. What are the team learned since then? Quite a

Andrew Lapthorne  41:23

lot, actually. I mean, George, and I build a lot of our stock screens, and we were sitting around with all these factors. And without being slightly disingenuous to the quant fraternity. It’s a bit like we’re building a cake right used to have a bit of value have a bit of momentum, or quality? How should we define it? Put that in there as well. Let’s run all these back tests. How’s that look? Or maybe Chuck something else in? I mean, we, we have all these factors. And we optimize these factors with the choice of factors we choose in the first place and combination of factors. Do we do earnings momentum plus value? Do we do price momentum, then do we start with quality first, and you’ve got lots of different combinations of ideas, which we try to put into a logical portfolio. Now, George is particularly good at the factors he’s very experienced at the factor library. And he didn’t really have any experience of machine learning. My machine learning expertise came from my university degree, when actually calculating anything was rather difficult. And the lights tended to go dark when you’ve got the calculations running. But the technology has improved so much the accessibility of really complex algorithms, you don’t need to build them. They’re they’re available to you. And obviously, the computing power as well, is extensive now. So we’ve got this complicated data set, we know what we’re trying to do, which is forecast essentially one month ahead, returns or maybe 12. month ahead returns. So we had the data. And we knew what we’re going to do. So we just said, look, let’s get the machine learning system to have a look at it. And, yes, support vector machines is better than this and better net, but actually, they all tend to give the same result, because all they’re really doing is optimizing factor performance. But what was really interesting and maybe kind of reassuring, it came to the same conclusions. So reversals is the most significant factor. But the most expensive to trade value factors featured a lot. In fact, free cash flow based EBIT, da factors features, which again, we’ve been doing the back testing, almost semi manually for decades, you can’t change history. So it’s not unusual for a machine came up with the same conclusions. But what I was really fascinated about is the combinations of factors, I kind of disregarded earnings momentum as a factor many years ago, because I just felt that it was being too positioned by corporates, the IR department moves the consensus around depending on whether they want to surprise or not. So in 1996, pretty much coinciding with the Ibis putting it out on a disk, it stopped working, because everybody, you know, the corporates knew it, people were focused on it. So as a factor, it was kind of ranking in the low 60s In terms of importance. And when we started putting it into machine learning system, it was consistently ranking in the kind of Top 10 Top 15, but not on its own in combination with other factors. Taking, you know, distance to default distance to default is a nonlinear factor. No one cares about balance sheet risk until they do and when they do, it’s the only thing they care about. So most of the time, having a low beta versus the market is meant to be a drag on performance. So ya know, we’ve learned an awful lot. It’s confirmed a lot of things that we fought about the data, bad data and machine learning is toxic. We just experimented at giving it a look forward bias of today. case. So we misaligned the data, the machine picks it up very quickly it goes to the moon. So if you’ve got a data set, and you don’t know the date of that data set, if you don’t know where that data set has been played around with historically, these optimization processes are very good at picking that up. So it’s useful that we all know about factors. And therefore we know when factors are working when we’re not. So we could interpret the data, because we’ve experienced those markets. But if you take in a brand new data set, I don’t know, satellite data, or car park data, we don’t know what it’s doing, because we haven’t experienced it before. So it’s very difficult to work out whether the model is doing things correctly, and therefore you’re at total mercy of the black box. What we’re doing here is we’re taking something which we’re already doing, and we’re just applying a better optimization engine. And I have to say, in terms of style timing, it beats the hell out of me. I mean, it was unloading value very, very quickly. And funnily enough, it’s just, it’s fully loaded back in. So you’ve essentially created yourself a factor trend following system is working out what’s important. And even though we’re using very long datasets to build the model, it’s quite reactive. It’s surprisingly reactive.

Corey Hoffstein  46:19

One of the things that quants usually need, or historically quants have needed is a long depth of data in terms of being able to do their research. And it makes it arguably difficult when there are new emergent phenomenon that are coming to the market. One of those over the last decade and very present in 2020 has been the growing influence of monetary and fiscal policy. I’m wondering what your thoughts are on how those will impact the traditional risk premia research landscape,

Andrew Lapthorne  46:53

I mean, even a lot of data which we expect to be correct, I remember looking in one particular database for German stocks for DAX Vertie. And that database had free stocks in the database. If I go back to 1996. And look at that database, it’s called 30 stocks. So I would have needed to go upstairs and type in all the data from reporting accounts. So there’s a lot of biases in the data that we see. But the central bank policy, I think, it’s always been there, you’ve always seen interest rate changes, you’ve always seen monetary policy adapting to the cycle. And both cycles have, you’ve had downside risk, which is really just been related to the cycle. What has really changed over the last seven years? Well post the financial crisis really is how almost global economies cannot cope with asset markets going down. And I think that is changed how people perceive certain assets. And it’s also changed how certain assets perform. So let’s take value versus quality. It’s been systematically D rated since 2013. And its economic performance. So the profitability of those value companies has not been that far away from the profitability of equality. And I think we wrote a piece which shows that you could explain pretty much all of values underperformance over the last five or six years through a derating of the stocks. Now, value suffers from rewritings and de ratings. In fact, most of return is coming from a rewriting. But the fact that was doing this during a bull market was fairly exceptional. And as I said, as I pointed out before, to have a market which has been predominantly driven by a rewriting of certain assets and a derating of cyclical assets, to have a bull market where value stocks were going backwards is very unusual. And to have a market which is going up when profits are stagnating is very unusual. And I think that’s been the key differentiating factor, I create a chart. Quite a few people have seen it where I just split the universe, whether it’s the s&p 500 or MSCI World or stock 600 into quintiles based on bond correlation. So your quintiles now our five year bond correlation, and you could change your treasury to your bond, it doesn’t really make much difference. And typically, the valuation on those two assets has been reasonably the same. For the last 40 years. It diverged massively from 2013 onwards to the extent that your quality stocks are trading on a price to book of for poor quality growth stocks, and your cyclical stocks are trading on a price to book of one. And for me, that divergence is a function of what’s going on with monetary policy. How quickly they revert is obviously a key topic. So I think they’ve had a huge impact, but central bankers still can’t get bit of the cycle. And I think that’ll be That’s the interesting element going forward.

Corey Hoffstein  50:04

One area where the US is still playing catch up to Europe is in ESG. Investing and Sri investing. I know this is an area where European quants have probably had almost a decade advantage on us quants. And looking into this type of research, What lessons do you think us quants should know, as we start to see es G gain further traction here?

Andrew Lapthorne  50:31

Oh, it’s gonna come at you fast, that’s for sure. That’s, I mean, there’s a huge transition going on, there’s a huge transition, I equate it to as big a transition as when the Euro came about when the Euro came about almost overnight, people switched from what were local currency, you know, French, German, Dutch portfolios to effectively Euro zone portfolios. And we’re seeing a huge amounts of money move from non ESG to ESG compliant benchmarks. And I often say, we have this roadmap, we have 15 different maps, some of them are going in completely the opposite direction. But there is a lot of this coming down the road, we’re already seeing it on a day to day basis, it’s not something that you could ignore, it’s coming from the asset owners, it’s coming from investors, investors want to invest in things, which are better for the planet, they want to have an influence. So it’s something that I think everybody every investor has to deal with. And we all know the challenges of data. You know, MIT wrote a very good paper a few years ago, just looking at the top five ESG providers on how there was a huge disagreement from those providers. Now, we like to do things which are quite precise cons like to do things which are quite precise. So when you’ve got, I think MIT basically said, we’re trying to measure people and try to measure funds and corporates on something that we’re not measuring particularly well. So how do we optimize towards an idea, which is still it’s been around for a long time, but the datasets are reasonably in their infancy. So I think it’s something that we whilst it’s full of challenges, I think actually, this is what quants are really good at. I mean, we’re really good at building portfolios, I think if you have an objective. So let’s say your objective is I want to invest in Japanese equities. Well, that’s subjective, you buy Japanese equities, you build a portfolio, which is appropriate for Japanese equities. Now, just because you want to invest in Japanese equities doesn’t mean that you’ve said I want to outperform, you’re just specifying that I want to invest. So a lot of this requirement is coming from something which is different to risk return. It’s a constraint, but doesn’t necessarily mean it has to impact overall performance. The other challenges with data set is bringing with it a set of biases, as you said earlier, ESG has probably had greater momentum in Europe and therefore ESG data, and the provision of ESG data from corporates is probably better, that will give you a global portfolio a bias towards European equity, and you would end up underweight US equity. And then there’s also the sector risks and the exclusion. So often the performance is relating to the biases in the data set, not the message that is coming in terms of are you good or bad ESG company? Handily, as I said, once we correct for that, once you correct for those biases, actually, the data becomes I wouldn’t say irrelevant, but I would say the target of ESG compliance is less challenging. And actually this year, for example, our pure factor ESG index is performing broadly in line with its target benchmark, despite the fact you’ve seen strong rallies and things like you know, commodity sectors, and energy intensive sectors, which are typically Well, low carbon in particularly unfriendly. So yeah, I think it’s a huge, huge challenge is 1000s and 1000s of data items to look at, but there’s no avoiding it. from a commercial point of view. This is what a lot of people are demanding.

Corey Hoffstein  54:18

Last question for you, as vaccine rollouts are happening around the world, and I’m keeping my fingers crossed that these new strains and variations don’t cause any hiccups. Assuming the world is hopefully starting to get back to normal in the summer and fall. What are you most looking forward to?

Andrew Lapthorne  54:37

Well, I was hoping to go to the pub on Monday, but typically, events snowed. I mean, you know, middle of April, we’ve got like an inch of snow. Now I’m definitely interested. You know, meeting clients is what I’ve been doing for a long, long time. But you know, the conversations with clients are what makes the job fascinating. That dialogue, that sharing of ideas I think it’s going to be a while before we’re getting on planes and going around and meeting people. But at least being able to wander down the road and have a drink with a client and have a chat about markets, I think is one of the things kind of workwise. I’m looking forward to a bit late in the day, I would have loved to go skiing, but I think I missed that particular season. So I’d like to get back to the Alps, and see the mountains as soon as possible.

Corey Hoffstein  55:25

Well, Andrew, I can’t thank you enough for joining me, this has been a fantastic conversation.

Andrew Lapthorne  55:29

Thanks very much. Thanks for having me.

Corey Hoffstein  55:35

If you’re enjoying the season, please consider heading over to your favorite podcast platform and leaving us a rating or review and sharing us with friends or on social media. It helps new people find us and helps us grow. Finally, if you’d like to learn more about newfound research, our investment mandates mutual funds or associated ETFs, please visit think And now welcome back to my ongoing conversation with Harley Bassman from where we stand today, in 2021, how do you think about the role of diversification as a primary means of risk management for investors going forward versus long convexity positions?

Harley Bassman  56:18

I don’t think it’s versus I think convexity is what you can add or subtract to the core asset you own. Okay, so you can buy a stock or you could buy a call and stock where you can sell a put on a stock or you could do any of these things as ways to capture exposure to the core asset. It’s not or it’s and now, diversification becomes interesting question. And this actually, I would say is my greatest fear what keeps me up at night, you’re kind of stumbling over the truth here. What we’ve seen in the last decade is that stocks and bonds go in opposite directions, least locally, stocks up bonds down and vice versa. They act as a quasi local hedge for each other. And thus, Bridgewater has become a billionaire. Because what did he do? He said, Well, let’s go and take $100 of cash, we’re going to invest $130 in bonds, and 70 in stocks. And they got that waiting by looking at that look correlations. So they’ve taken his $100 levered it up to one at a ratio where the two are somewhat offsetting. And that’s worked very well. And I say congratulations. It was an incredible observation, and a brilliant execution. So he deserves to be a billionaire. Despite some of the other things a little crazy about it. The problem we have is a weakened own that correlation. Will we realize that correlation forever? The answer is no. As a matter of fact, this correlation is a relatively recent phenomenon. And prior to 2000, more often than not you had stocks or bonds with the opposite correlation where they went up and down together, it is my belief that we’re going to have higher rates, which we could talk about a different time. And it is a I will say a fact but a historical artifact that when you see interest rates and inflation get above three and a half, four and a half percent somewhere in there, the correlation flips, if it is the case that we get interest to rise over the next few years. And we have simplified products to go and solve that problem for you. And we see the correlation flip, it’s gonna be a real problem, because all the guys who are in the risk parity trade where they own stocks, they own bonds, and levered basis, they’re gonna be sorry, a lot. People who were using simple diversification portfolio like 6040 strategies with stocks and bonds hedge each other will be sad also, because they’ll go down together, it won’t be as bad because they’re not levered. But they’re both going to go south. And if you look back in last five years, we’ve had two or three deep drawdowns, where stocks and bonds have gone down together, where it’s basically correlation, one, everything goes down. Well, that kind of came back with the help from the Fed. If we get rates above a certain level, the Fed will not be able to go and stop that correlation one. And so when you’re talking about diversification, usually thinking a bit broader than just I’m going to own some defensive stocks and some aggressive stocks as a fang stocks. When things get rough. There’ll be all stocks go self to some faster than others. I would be cautious now. And I might think of diversification as being a much bigger topic as my house like a house with a car. Do I have cash for six months of wages? If I get fired? I would think much more than just my personal trading account.