Adam: Let's start with the origins of Pecora.
Aaron: Ferdinand Pecora was a Sicilian immigrant to New York over a century ago, who against incredible odds, actually went on to get a law degree at a time when Italian-Americans just weren't able to get into law school. He was an American Dream/pick-yourself-up-by-the-bootstraps type of person who went on to make a very meaningful impact on the financial world. After the market crash of 1929, he went up against Charles Mitchell of National City Bank (now Citibank), the JP Morgan dynasty, and a number of other very powerful banking interests which had profited from the collapse. He went up against the most powerful men in the world on the basis of his integrity and what was right.
In the wake of this activity, he laid the basis for the Securities Act of 1933, the Investment Company Act of 1940, and all of the most important laws and regulations that are still governing the financial markets today. He then became one of the first commissioners at the SEC. So a relatively obscure person in terms of financial history but someone who actually had a pretty big impact. We find that to be a source of inspiration.
Adam: It sounds like you see yourselves almost in a kind of David-and-Goliath scenario here. Or is that stretching it?
Aaron: King David is one of my greatest heroes. Mighty Goliath taunted the Israelite army for 40 days before little David, the youngest of his brothers, courageously won the battle singlehandedly. Despite the odds, David was successful because of his faith and, subsequently, became the greatest king in the history of Israel. Therefore, faith is the basis of everything I do, both personally and professionally. When it comes to trading, the analogy is apt because the FX space is dominated by the status quo, the large banks. At Pecora, the idea of independence and integrity in our research is paramount. We strive to build an environment of autonomous idea generation without respect to quarterly earnings estimate pressures or any other constraints. We want to go where the data leads us. And we want to be objective in building the best trading technology possible, without any hubris, without any politics and with a clear sense of independence that enables us to allocate our capital to high quality trading systems. We won't fall in love with our own ideas -- the best managers adapt proactively over time.
Adam: That leads me straight into the analytical approach you take. What are some of your current ideas and how do you arrive at them?
Aaron: First of all, we're very focused on the development of short-term trading systems in the FX space and in liquid futures markets. We're not high-frequency. We're not medium to long-term. We want to focus on the sweet spot. In the high-frequency space, I think you're in a technological footrace. In the medium to long-term space, you're really at the whim of the general market and you generally show higher volatility and drawdowns, which are untenable, at least from the risk appetite of most investors today. For us, the sweet spot is short-term trading systems. We don't subscribe to any Holy Grail theory. We just focus on looking at different types of trading strategies and blending them together so that we generate the highest Sharpe ratio possible. Our core trading strategy dates back to 2008.This trading strategy, I would call it a momentum strategy, is essentially trading breakouts on a technical basis. This is, I would say, our trend component, which is one module or basket of correlation.
Another module would be what I would call counter-trend or mean reversion, and there we have a systematic approach. It is essentially selling intra-day highs and buying intra-day lows based on patterns that the system is able to recognise. This is actually quite different from module one. Module two has a completely opposite impetus for trade generation and a completely different approach which means module one and module two really have a quite low correlation. When you add them together, the risk return characteristics tend to improve.
Adam: You tend to get a lot of people falling in one camp or the other, mean reversion versus momentum. Is there not an issue where you are cancelling each other out a little bit? If it's in the same market would you potentially be taking different sides of the same trade?
Aaron: Yes and no. There are times when there are two different strategies -- a momentum strategy, which is long-euro and a reversal strategy, which is short-euro -- and therefore your net exposure is zero. This, for me, is fine because it means that open risk is low and volatility is low. But basically it's all also a question of timeframes. If your momentum strategy has holding periods, on average, of one to two days and your reversal strategy has holding periods, which might be, let's say, two hours. Imagine that the euro is trending upwards, from 137 approaching 140. Over the course of five days, the momentum strategy would have a long position, and it could be that in the course of that specific trading session you will see an intra-day high and an intra-day low. So the reversal system might be able to pick off intra-day reversals inside of the overall trend.
Adam: Going back, you talked about pattern recognition. How much of that is automated or is systematic?
Aaron: Our thinking is really systematic, so to the extent that we can be systematic, we are. I think the human touch really comes in outside of those components. In module one and module two we have a database of 1 billion ticks of data that we can back-test and run on a daily basis. There's no reason those strategies shouldn't be 100% automated. There's just no reason for it. Where the human touch comes in is if we go to module three. Module three would be what I would call discretionary macro-trading. Essentially this is the skill and ability of the portfolio manager, who hasdeveloped over the years a certain savoir faire for the FX markets, where he says, "You know what, based on the technical picture that I see, based on the feedback I'm getting from other market participants etc...." This is something that we don't rely heavily on, but do think of as a third strategy, which also doesn't correlate to mean reversion or trend following. We believe that there's a diversification benefit and a residual benefit from the fact that this macro-trading will correlate to the other positions.
What typically happens is when we have a thesis or an investment idea, typically the portfolio manager will actually look at the portfolio. If the macro-trade would increase the open risk,we would not likely take the position. If the idea is to buy euros and we already had a net long position in euros, we would not take the trade. If it is a trading idea to short the euro and we had a net long position in euro, we may do it in the sense that it's actually reducing the net open risk.
Then on top of that, we have a discretionary overlay because the question that we're asking is, "How much do you allocate to trend following, how much do you allocate to mean reversion, how much do you allocate to discretionary macro trading?" Well, it's not a static number. What happens in practice is we evaluate the trading environment. For example, the volatility in FX G10 right now is extremely low. That's a bearish indicator for trend following. We don't want huge exposure in trend when the vol is at multi-year lows.I think this dynamic overlay is a very important aspect of what we do. We can increase or decrease our allocations from different buckets based on the trading environment that we see.
The other matter is over the past five years, we've seen over 500 coordinated global central bank rate cuts. It's the most coordinated activity in the history of mankind. Look into all of human history; nothing has been as coordinated on a global scale as the past five years of central bank policy. Every single bank around the world: cutting, cutting, cutting, 500 plus times. In this environment there is no meaningful interest rate differential in the G10and the traditional driver of currency prices is interest rate difference, so the main driver for an entire asset class has been eradicated.This is why most FX strategies have struggled in the past few years, but that doesn't mean there's no driver. It just means there is a secondary driver, which has now become primary, and that is the news surrounding monetary policy announcements of various central banks. Now we have a trading environment which is materially different from years past. You go into a Fed announcement and you have a plus one or a minus one impact to your portfolio instantly, what the media calls risk-on, risk-off. You're either a hero or you're a zero, and that's not something that we have an interest in in our portfolio. What we'd rather do is be intelligent about trading around these events.
To some extent, that's systematic. For example, in our momentum systems, we don't trade on NFP days when non-farm payroll comes out. That system's just flat that day.
A portfolio manager in this case is able to assess the portfolio going into these high-impact news events and assess whether he wants to be flat, whether we want to hedge our position, or whether we're totally comfortable and we want to stay with our position as it is. I think this discretionary ability of the portfolio manager to hedge-risk, or to frankly just go flat and close all positions, is something that's very unique as a quant fund. Most quants don't do that.
The reason why most systematic funds have lost money overwhelmingly over the past few years is because they go into trading announcements and get killed. I'm not saying we've cracked the code and have a perfect way to handle it. I'm just saying that sometimes we take our risk off the table. You know, maybe we would have made a lot of good money and maybe we would have lost a lot of good money. I'm fine with that. I'm always fine with taking less risk.Our perspective is, we are the main investor in Pecora's strategy. We're the ones putting our money in the trading systems. It's more important for us to stay within our volatility and drawdown budget than to hit a certain type of performance in a particular year.
Adam: There's been so much discussion about how the ultra loose monetary policy is going to change and when that's going to change. We've already had the initial signals of the tapering discussion but it doesn't feel at the moment that the FX world has changed dramatically. How do you see things going forward? I'm talking about when the overall paradigm has shifted and we are in a more historically normal period, as it relates to politics and FX behaviour. How will you be approaching that?
Aaron: That's an interesting question. It begs the question, what is normal? The markets today are different. I wouldn't say that they're abnormal. Over the past century, short-term interest rates have been 4% on average. So yes, it's kind of strange to have zero interest rates. If you look at any time over the last century, it's pretty rare. There are really only two scenarios. One is you stay in an indefinitely low interest rate world, for an incredibly long period of time.
Adam: A "new normal".
Aaron: Yes, it's a bit like the new normal. The other possibility is that monetary easing and quantitative easing end and you see a significant correction in asset prices.
As I see the macro-environment, the S&P 500 has a 90% correlation to the Fed's balance sheet. As long as the Fed balance sheet expands, equity markets go up and that's just how it works. They're 90% correlated. The policy maker has two choices and the market, for the record, is going up for no reason other than policy. There is no real economic basis for PE ratios being 30% higher than their long-term average. There have only been two times in history with PE ratios this high, and that was in 1929 and in 1998-99 tech bubble. Only in those two cases in the past century have PE ratios been that high. Margin interest on stock purchases is at an all-time high. In that context you think to yourself, the policy makers have two choices: Either they temporarily let some air out; they induce a correction so that things don't overheat. How do they do that? Stop tapering, stop printing for a little bit. That means the market goes down, but hopefully in a controlled way.
The other possibility is the policymaker says, 'We can't stop this, the economy's too fragile. We've got to keep printing." Then you have a prolonged low-interest rate environment, in which case the market itself will eventually break under the weight of the monetary stimulus.
How exactly is the environment any different right now than it was under the bubble that was created during Greenspan's chairmanship? The only difference is GDP growth is much lower.
In any of those scenarios, I think it's actually quite interesting for our trading systems to continue if there's a momentum strategy, because they always tend to do well in these crisis periods. A point in fact would be 2008. Our trading strategy did very, very handsomely. It's no secret that momentum strategies can do well when volatility's expanding, when you have crisis periods. That's not something that we're afraid of for our portfolio, although it's kind of frightening because we're looking at this very slow-motion train wreck, and one would hope that the economy catches up to the policy.
There's this sort of fake economic expansion based on monetary policy. Quality, value, and innovation are what we need to see in the market to have a sustainable positive scenario where middle-income, middle-class people have a better way of life. Not just multi-billion dollar corporations and high net-worth investors, whose paper assets are going up because they can borrow money infinitely at no cost to themselves. As an average person, if you go in, start a small business and try to get a loan from a bank, they will escort you out of the premises.
Adam: That ties back in with what we were talking about, getting back to some kind of historical norm in terms of the relationship between policy and economic performance. Do you think that the systematic approaches you take will adapt? Or will that be a time to say, 'Okay, the world has changed. Now we've got to change tactics entirely'?
Aaron: Here's my thinking: we don't want to be rigid in our view. We want to be objective and react to what is, not what we think will be, but what is. We have to follow the data wherever it leads, whether we like what we see or whether we don't like what we see. Systematic trading will always have a basis and will always 'work'. I mean, it's almost infinite. It's unconscionable for me to say, 'Oh, systematic trading doesn't work.' Even within our niche, which is short-term trading systems, there are so many permutations of different trading ideas, there are so many possibilities. I think that we're going to be able to develop new systems on a continuous basis and add them to improve the risk internally. It is a tricky environment. We made money last year. We made around 7.7% -- that's half of what we expect to make in a typical year, but we can live with that. We just want to stay within our volatility and drawdown budget.
If we made zero, I'm cool with that. If we make 20, I'm cool with that too, as long as we stay within what the principles of what we're doing and budget our risk accordingly. We have very, very long-term thinking. We're building systems with a 20- to 30-year view, but we're not static. I like to think of us as an evolving CTA. We have core trading systems that are tried and tested and proven. But we believe that in technology, we want to evolve and continuously use our research methodology to produce new strategies that can add value.
Adam: From a technology point of view, are there any specific things you try to do to complement that philosophy? How big a part is technology in what you're doing?
Aaron: In some ways, technology is everything. In some ways, methodology is everything, because we have great computational capability. We're dealing with big data. Ideally we have the capacity to run 100 million back tests in a year. We're talking about extremely large numbers. When we develop a new trading system and bring it online in real production, it's not like finding a needle in a haystack. It's like finding a needle in 100 million haystacks to get that one strategy, so we're dealing with incredibly large numbers and brute force computation. That being said, it's really a question of methodology, because if you don't have the right methodology you could spend the next million years testing the wrong stuff and that's something that's proprietary to our research. It's how we test things and how we define a system as being robust or a strategy as being robust. Our own proprietary definition of robustness I think is critical, because if you did that brute force computation without the methodology behind it you'd get nowhere.
Adam: What are you finding from your business point of view in terms of technology trends? Given the size of the data you work with and the amount of calculations you're running, are technology costs and trends a performance issue for you?
Aaron: Good things aren't cheap. Quite a large amount of our costs on a percentage basis are technology related. I would say much higher on a percentage basis than a typical hedge fund. Technology costs come down and are coming down -- XQL databases and stuff like that. Our access is getting cheaper and cheaper all the time. That said, our principle is we spare no expense when it comes to technology and we spare every expense when it comes to everything else. We have an extremely lean business model where we're extremely circumspect about every penny we spend, but on the other hand, we're in a world where we're really quite liberal and generous with our technology budget.