Nobel Laureate Dr. Myron Scholes discusses the Adaptive Multi-Asset Solutions Team’s approach to investing, with a focus on mitigating risk of significant losses while seeking total return.

Transcript

There’s a debate among advisors and among investors as to whether active management or passive management is the right place to put money. So, that has come about in part because of the great growth in ETFs and in part because of the average inability of active fund managers to outperform benchmarks.

I think that the debate, however, is centered on the wrong question. The debate is centered on performance measurement. The debate is centered on whether a fund manager can outperform a benchmark, and a lot of that performance benchmarking constrains the managers to stay close to the benchmark. For example, Morningstar will rate a manager lowly if they deviate too far from the benchmark. And the question as an investor, the investor should raise, is not whether they’re close to the benchmark, or how well they do relative to the benchmark, because that relative performance, it doesn’t really bring home the bacon in the sense that it’s absolute performance that has the most important effect on an investor.

My view of investment has been that there’s a right way to try to generate alpha and an incorrect way to generate alpha. My thinking is that if you have, if you think about investment as this relative performance to a benchmark, and you’re too close to the benchmark, then you’ll stay very short term in your focus to generate alpha. If, on the other hand, because of the constraints of so many investors to stay close to the benchmark, the active investor has a theme, an underlying theme that they understand, people who are constrained can’t use that theme, then they can outperform. So, really, in investment management it’s the constraints, the constraints of others that allow you to make excess return. So if an investor is, or many investors are stuck close to the herd, close to the benchmark, and because the active manager has the trust of his clients, he can, or she can then deviate from the benchmark and take tracking error, and they have faith in their long-dated themes, they can make excess returns.

The interesting thing, if you look at the very popular index fund, it has no tracking error, so that’s great, it’s highly constrained to be by itself. Two, it has very low fees, no fees or very low fees. But it has no risk management, so the risk of the S&P 500 in 2007 and 2008 was much different from the risk of the S&P 500 in this last six to 12 months, and obviously in 2012 the risk of the S&P 500 was much greater in the first few months than it is now. So the risk is always changing.

A story that I like to tell people is the movie The Titanic. Everyone in the upper decks of the Titanic were drinking and having a good time and dancing and listening to music, while those on the lower decks of the Titanic obviously were suffering and dying. So, relatively, those people on the upper deck of the Titanic actually were much better off than those on the lower deck of the Titanic. But as this picture shows the most important thing is not relative performance or relatively where you are, but absolutely they were not in very good shape. So the fundamental question if you’re constrained to stay at the benchmark, you don’t ignore risk but if you’re not constrained and you will trade tracking error, you can earn rewards by deviating from the benchmark.

The investor wants to maximize the compound return of their investment to get the best terminal wealth and they’re interested in how much drawdown they have to take to achieve that terminal wealth objective. The problem with indexing, or the problem with passive investments and/or active investments, in some sense, it’s all based on the jargon alpha, it’s all based on how well you do relative to the benchmark. The important thing about managing money is total risk, it’s not relative risk. It's how your portfolio is doing and what causes the risk of your portfolio to change, and if you don’t incorporate thinking about how risk changes, and how that will, risk changes will impact on the compound return, then investment management is only using one hand clapping, it’s not using both hands clapping. And my view is that risk management is much more important, much more important than deciding how much active or how much passive investment to hold, because at times of shock in the market when uncertainty increases it’s very hard to tell the difference between a passive investment manager and an active investment manager. They’re very, very highly correlated with each other. In managing money, the debate should be how to manage the risk of the benchmark.

If you’re just holding the index, you know it dropped 50% in 2007-08, and it dropped 50% or 55% in 2001, you know, and you could drop 20% or 15%. And if you’re interested in compound return and not average return, those hurt. Those big drawdowns hurt. And they hurt the investor because once you suffer a large drawdown it takes a long time to recover, you know. And the interesting part of investments is that it’s the large drawdowns, or experiencing a large drawdown or missing a large upside, that have the greatest effect on investors’ terminal wealth. Investors are interested in terminal wealth. They’re interested in how well can I accumulate, or what can I do over time because my terminal wealth gives me money to consume. It gives me money to consume to send my kids to school, or gives me money to consume to take a vacation, buy a house, retirement, medical, etc. And those are very important to investors. And we’re tending, if we concentrate on relative performance, we miss what the investor really wants. The investor really wants a growth of their portfolio but worried about how much downside I’m going to take. Why are you worried about downside? The investors are worried about downside because they might need to consume, or they don’t have money. Or they might want to retire sooner. If they take the drawdown they’re going to have to retire later in life, and those are very important considerations. It's the drawdown which is the most important risk.

And the interesting part about compound return is that every period matters. You can’t say as an investor that you’ll have a long horizon, like I’m going to be invested for 10 years, because if you start and you say you’re going to be invested for 10 years, if in the first year you lose 90%, it’s going to take a very long time to catch up, maybe never.

Investment is not averages. Investment is compounding. And what I mean by averages is, average returns assume that we’re always investing the same each period. So if we start off with 100, we invest 100 today, we make money, we take the winnings out, consume it. If we lose money we put that back in and we’re always investing 100 each period. Compound return, on the other hand, accumulates, it takes the money, the 100 an investor has today, and if they make 10% it’s now 110.

It’s always accumulation until such time as one wants to convert that to consumption. And so basic compound return, the mathematics of compound return, are much different from average return, because average is simple. You just take the sum and you divide by how many units you had, and you just divide, and you assume distributions.

If we look at the returns from 1857 to the end of 2016, we found that if one invested a dollar at the 1857, held their portfolio through to current time, that that one dollar would compound at about 5% a year. So one dollar would grow to over $3,000 over that period of time. But if one were fortunate and you sold all your stocks for the worst of the performing months over that period of time, then your compound return, instead of being 5%, would be 9%. So avoiding the worst months, you end up with a 9% return, which means instead of your dollar growing to $3,000, it grows to $1,400,000 over that period. And what, on the other hand, if you sell all your stocks off for the best months in the market, instead of your compound return being 5%, it would only be 1%.  One dollar grows to $8 over that period of time. What this illustrates to me is that the lion’s share of returns are determined not by the little ups and downs that you have each period of time, but by the tails of the distribution. In other words, participating in the gains is important, avoiding the losses are important. They have the largest effect.

For the period 1/1/1857 – 12/31/1925, individual security returns were gathered from U.S. financial periodicals on a monthly basis, beginning with the official list of the New York Stock Exchange during that time period. From the period 1/1/1926 – 12/31/16 returns are represented by the S&P 500 Index. Source: Ibbotson. Past performance is no guarantee of future results. Assumes reinvestment of income and no transaction costs or taxes. This data is for illustrative purposes only and not indicative of any investment. An investment cannot be made directly in an index. An extreme tail gain or loss is described as any monthly period whose performance is 2 standard deviations above or below the average monthly return for the entire period. 

Again, it’s not focusing on the average because the average tries to wash away the tails. It wants to trace what’s the average experience.  But, the average experience is not what we live. In our world we don’t live the averages, the tails affect everything we do. And if you look at the data, the number of big tails are much larger; the tail results are much larger than would be anticipated by a normal distribution. Many more observations that we had, many more monthly returns, or quarterly returns fall greater than what we predicted in the tails. So the number of bad outcomes far exceeds what you would expect in a normal distribution.

Diversification is important to investors most of the time, but the problem with diversification is that it fails at time of shock. Diversify away the idiosyncratic risk, get rid of it. Then what are you left with? You’re left with the systematic risk, or the beta risk of the portfolio.

And so when we think about our technology to use that information, it’s these different and conditional volatilities that we use to stitch together the conditional distribution to create an unconditional distribution. So we get the forward distribution of risk essentially by stitching together all the probabilities, the local probabilities, and adding them all together to get the unconditional distribution. So when all the volatilities are the same, that’s a normal distribution,

because the distribution is no different, you know, and the tails or the middle as the top. That when the market is telling us that the probability of loss is higher, so when probability of loss is higher, that distribution has a fatter tail to it, or more probability, or more density, and as you go into the money it has thinner tails. So since we like positive skewness and we don’t like negative skewness, if the expected tail gain is higher than the loss, that’s positive skewness, right?  So we like those assets. We don’t like the ones where have expected tail loss is high relative to expected tail gain, so the optimizer would move away from those assets to those assets that have high expected tail gain to expected tail loss.

I think they happen infrequently and in measurement as they said in the past, but I think that there are myriad little tail events that are happening, and there are idiosyncratic tail events. I mean, you know, it’s not necessarily the case that it’s just based on broad markets.

The problem is that you think that you might be more diversified than you are, even to idiosyncratic, to diversifiable risk, or more systematic risk, or market risk, and so you don’t put all your eggs in one basket. I worry sometimes, you know, you have eggs in this basket, you have another basket on your arm, and a third basket on your arm, and similarly on each of your other arm, you have six baskets, all not putting all your eggs in one basket. But then you’re walking with all of the eggs, and you trip and fall, all the eggs break.

Friends of mine told me, “Myron, I was stupid to have such a large home on the banks of the Mississippi,” when I was at the University of Chicago. So I sold my large home on the banks of the Mississippi and I bought a small home one side and another small home on the other side of the Mississippi, so I was diversified … I diversified my risk. But I was diversified to idiosyncratic things like you know fire, or whatever, but certainly I wasn’t diversified to floods. So diversification is good for the idiosyncratic risk, but the systematic risk part, everyone suffers.

60-40 strategy, which is a strategy that is a static allocation strategy, is an average strategy. Again, it doesn’t take account of the fact that sometimes you’ll have very negative returns between bonds and stock, and sometimes you’ll have very positively correlated returns between bonds and stock. And so sometimes the risk of a 60-40 strategy is 90% growth assets, 10% bond.  Sometimes it’s 80%, you know, sometimes it’s 30%. If the investor wants to risk manage their portfolio to keep it at target, then how can you be keeping at a target when sometimes your risk is equivalent to 90% stock and sometimes it’s equivalent to 30% stock? And so by adjusting the relative weights of equity and, or growth assets and bonds as opposed to keeping it static 60-40, will give you a better experience than just the static strategy. And so static strategies work if you don’t have information. You get average results, but average is average. We don’t want average results; we want superior results, superior compound results, superior compound return, superior terminal value. And if you use the market information to adjust the 60-40 strategy to move sometimes to only 30% in equities, and sometimes maybe 70% or 80% in equities, and achieve the same average risk level, or same target drawdown level but end up at a higher compound return, that’s a better strategy.

I believe that the credit markets, the option markets, have a lot of information, and they don’t have great information maybe for one-year options or five-year options, but they have a lot of information about near-term risks, three-month risk, two-month risks. Do market prices in the option market have information that can be used to determine the distribution of risks? And the answer to that is definitely yes. The option market is an insurance market. It’s a strange insurance market in that the put option protects your downside, and the call option is, you know, protects the upside.

The people who are, are the very active and smart traders who realize that risk is increasing and they might be stuck at the assets they’re holding, they go with the option market because the assets they’re holding might be illiquid but they want to buy market protection. They want to protect against the beta risk and so they reveal themselves through the decision to implement the protection, or to lift the protection, or reduce the price of the protection, or to participate in the upside, you know, through acquiring out of the money calls.

The way I think about it is that the option market should have a lot of information … and it does. If you look historically back even to 1987, the out of the money puts were increasing in value before the market crash of ’87. And similarly you had the case in 2001, and you had the case again in 2007 and 2008, the option markets were, the prices of puts were increasing dramatically before the actual market went down in extreme ways. It was going down and then the price of the puts also indicated that the probability was higher that the market would continue down because people were buying protection.

So the information in the risk markets tells you how to adjust your risk of your portfolio. They’re all in the same loop. So now the question becomes why is it the case that the equity markets don’t incorporate this information, and they could. I mean they could. They could, it would be my joy in life to see that the equity markets reflect the risk dimension correctly. And I do believe the reason we don’t see it, however, and it’s my passion in life to get us to see it, is because of this constraint, because of the tracking error constraint. The indices you have, whether it’s the MSCI, ACWI indexes, or the S&P 500 or all the passive investment funds, they’re all relative performance, they’re relative values so everyone sticks to the benchmark. Everyone’s afraid to deviate from the benchmark. If you say the cost is my job to deviate from the benchmark, if the benchmark goes down 50% and I’m at the benchmark, I can be in the herd and everything is fine, I save my job. But is that the job one really should have? The job should be to maximize the terminal value of the portfolio for investors. So if that’s your job, you should be compensated that way, but the job that many managers that I talk to, the job that many allocators that I talk to have, is to outperform or stay close to the benchmark.

The economic rationale is it’s very costly to generate the trust of investors, and it’s very costly to measure their performance if you don’t have a benchmark. Trading off that cost, in lost return because you’re not using information to deviate from the benchmark, leads to an implicit cost in terms of lost return. So you have an explicit cost you reduce, and implicit cost you increase because you’re not dynamically managing the risk of your portfolio to enhance compound returns.

In using option prices to estimate the distribution of risk that the market is forecasting, the prices have to be good prices. And you need liquid markets. The prices we use to estimate the distributions themselves are the shorter end of the market. You know, the two-month, three-month options that have a large open interest, have very narrow spreads, have a great amount of trading volume so that market information prices are rich in information. And so that’s number one. Number two is since there’s a great cross-section of information in the option market, there’s not just one strike price. The whole, the distribution of prices, and one could use that information not only in the one strike price but of all the prices of options for that particular asset to be able to then figure out whether that sequence, or the whole series is a good sequence to use. You know, one of the nice things about tail risk management is in the growth sector, if one asset has a bad result, other assets are going to have bad results as well. So, the information is rich in the cross-section because you have a whole cross-section of growth securities you can use to estimate.

Janus Henderson offers active management products. We are complementary to so many of the other Janus Henderson strategies because they’re focusing on outperformance of a benchmark. And our job is, we’re beta managers, or risk managers, and they’re asset selectors, generally are trying to select assets that outperform.

A solutions business is different from a product focus. We have products, but at the same time we hope, we work with investors. Our job is to enhance the performance, or mitigate, enhance the compound return experience. I’m not saying the average; we’re in the compound return experience through dynamic risk management strategies.

For the period 1/1/1857 – 12/31/1925, individual security returns were gathered from U.S. financial periodicals on a monthly basis, beginning with the official list of the New York Stock Exchange during that time period. From the period 1/1/1926 – 12/31/16 returns are represented by the S&P 500 Index. Source: Ibbotson, based on monthly returns.

Assumes reinvestment of income and no transaction costs or taxes.

An extreme tail gain or loss is described as any monthly period whose performance is 2 standard deviations above or below the average monthly return for the entire period.

Options (calls and puts) involve risks. Option trading can be speculative in nature and carries a substantial risk of loss.