by David Harding on 19 October 2016
Estimated reading time: 8 minutes
Over the last two or three years, a number of new, interrelated fashions have emerged in the institutional asset management world. Smart Beta, Alternative Beta, Crisis Risk Offset, Tail Risk Protection, Positive Convexity have all arisen out of the defining challenges for asset managers today: risk free rates of zero, government bond yields in all major countries that range from negative to low single digit territory, and equity markets that have risen for years since their post financial crash lows and have for some time seemed very fully priced. These fashions build on the undeniable notion that diversification improves overall portfolio stability and performance, and therefore seek to raise portfolio risk-adjusted rates of return by increasing allocations to less correlated asset classes.
What is labelled as an “asset class” is something of a moveable feast these days. Real estate, forestry, private equity, junk bonds, hedge funds; all are variously described in such a way. Significant intellectual battles are fought over the use of such terminology. How correlation should be measured between asset classes is also a vexed question. Not all “asset classes” are equally liquid and correlations measured over different time horizons will present radically different pictures of appropriate asset allocation. The general trend induced by the current challenges however will be a favourable disposition towards assets that are able to present themselves as having a low correlation with those that traditionally inhabit investor portfolios, such as stocks and bonds. These trends have gained considerable respectability and momentum over the last 25 years. Pioneered by David Swensen through the Yale Endowment Model, the adoption of substantial allocations to private equity, real assets, and hedge funds has become the norm throughout the professional asset management industry.
One sector to benefit considerably from this has been the managed futures industry. In the 1980s when I entered this business, it was essentially an adjunct of the commodity futures brokerage business. Speculators, attracted to the high leverage available in the futures industry, hired third party money managers (CTAs) to trade for them. Some of these money managers came straight from the pits, others traded by the seat of their pants, but a not insignificant minority were tech-heads of one sort or another, with computer- based systems. Considered somewhat eccentric at the time, the tech-heads would have the last laugh, as their data-based, empirical, usually trend-following, trading methods worked well enough in the ensuing decades to make many of them rich. My explanation of this fact is that the revival of the faith in markets in the 70s and 80s (to Milton Friedman and Arthur Laffer, a son, Ronald Reagan and a daughter, Margaret Thatcher) went too far in anointing the Efficient Market Theory (EMT) as the formal orthodoxy of the newly emerging global empire of liberal, democratic capitalism. This rather arcane, mathematical doctrine out of the University of Chicago, held that markets, rather than being merely a “good thing”, were literally perfect. With the commanding heights of global capitalism occupied by EMT true believers, the opportunity for the “data geeks” to profit by “betting against those who don’t believe there is any point in looking at your cards” (as Warren Buffett says of EMT disciples), was considerable.
On the back of these developments the managed futures industry grew from a backwater to a significant slice of a global hedge fund industry, which is now, in itself, a major force in institutional asset management. Two other factors played a part in the industry’s rise. One was the enlightened set of regulations passed by the US Congress in the 1970s that created a regulatory structure which enforced high ethical standards upon what had previously been a pretty scurrilous industry. The second was that the futures industry grew beyond its traditional roots in the commodities business and into the far larger arena of money. By the late 70s and early 80s, foreign exchange futures, bond and interest rate futures and stock index futures were booming and, in terms of turnover, beginning to dwarf the commodity futures markets. What might have remained a fringe area of world finance, the Commodity Trading Advisory business, began to grow towards the mainstream.
By the late 1980s more “Commodity Trading” was taking place in stocks and bonds than in copper or soybeans. It was thus as big a shock to the tiny CTA industry when the 1987 stock market crash happened, as it was to the wider market. The event produced the usual cast of winners and losers, together with the requisite amount of breast beating. When the dust settled, the proximate cause of the instability was identified as the activity of “portfolio insurance”, an apparently sophisticated strategy that attempted to insure against stock market declines by actively selling futures into them. In this way, institutions had contracted to dump billions of dollars of stock in a plummeting market and this temporarily overwhelmed the supply of bargain hunters who were, presumably, temporarily shocked into inaction. Portfolio insurance was discredited, some circuit breakers introduced in the futures markets, but otherwise the system was left to heal itself, which it duly did.
Over the thirty years since then, the Iron Curtain has fallen, market capitalism has flourished, and then been disgraced by the global banking collapse in 2008. The volume of global savings has soared, as have stock market values, and levels of global debt. Longevity and living standards have increased worldwide, and to look after people during their long retirements, huge pools of capital have been accumulated. The managed futures industry has grown into a leviathan, now managing hundreds of billions of capital. Futures trading enables its followers to profit from both rises and falls in markets, and over the years that is exactly what it has done. The industry did well during the long equity bear market of the early 2000s, but it finally achieved glory (of a rather dubious sort) in 2008 when managed futures funds soared, both by riding the downtrend in the equity markets, and by being very long of fixed interest futures as they soared at the same time. In the wake of 2008, even the most CTA-sceptic of market participants had a hard time resisting their claim to a role in portfolios. Since 2008, the managed futures industry has grown considerably further.
Remember, however, that the industry’s very existence depended on rejection of the orthodoxy of market efficiency. Disconcerting as the events of 2008 were, to managed futures heretics there was a certain satisfaction to seeing their intellectual opponents (and oftentimes tormentors) put to the sword, their smug intellectual certainties tested to destruction by events. The “crowd behaviour” that is so evident in the study of price data (but absent in the idealised models of the EMT orthodoxy), was present for all to see in dramatic fashion. Had the authorities not intervened, it seems likely that even the vindicated techno-geeks (me, that is!) of the managed futures industry would have been denied their satisfaction. It would have been a Pyrrhic victory, as the entire mechanism of the modern financial markets could well have ceased to function.
A side effect of the rescue was that the EMT proponents, having looked into the abyss, lived to fight another day. Instead of the theory being totally discredited, it has survived, with amendments, but still recognisably intact. Its Greek letters live on, beta for the supposed skill-less return automatically provided in exchange for risk taking, and alpha, for the nectar of the Gods that is the reward to the elect for completing the job of making the market perfectly efficient. The assumptions remain: that standard deviation of returns measures (or even is) risk; that correlations between assets are meaningful long run measures of something; and that market “fundamentals” will yield up the absolutely correct market price at all times. Legions of mathematicians, economists and physicists have worked to refine the EMT models by adding extra parameters to explain away their practical failure. Some of these parameters are held to represent additional alternative risks (or betas) to the dominant “market” risk, among them value, size and latterly momentum. After 30 years of denying the possibility of momentum as a phenomenon existing in markets, the orthodoxy has now incorporated it into EMT with the status of eternal truth.
What is worse is that the orthodoxy can now investigate the properties of this absolute truth with the flawed statistical methods used in this quasi-science. It can measure properties such as its correlation with other factors and use terms such as “convexity” to describe hypothesised relationships as though they too were eternal truths. The bottom line is this: historic simulations of momentum-based systems look sufficient to explain the historic success of managed futures funds. They have done very well over the last 30 years and produced an excellent complement to traditional asset classes, for example by remaining stable or increasing in value in the two great equity bear markets of the noughties. Institutional investors are seeking returns and are nervous about the prospects for traditional asset classes. They naturally don’t want to pay more fees than necessary, so they are willing to invest large sums, on the basis of this weak theoretical underpinning. Many new products have been launched by firms inexperienced in this domain, often on the basis of simulated track records (or academic papers containing evidence of “alternative risk premia”, which amount to the same thing).
The activity is gussied up under a fancy but incomprehensible title like alternative beta, crisis risk offset, or tail risk protection. “Assets” that are favoured are those that can be represented as possessing the fabled property of “convexity”. This means that their returns are non-linear and can be expected to be positive when those of stocks are large and negative. What a burden of expectation is being placed upon the small supporting pillar of data. Correlation between stocks, bonds and managed futures funds is not a meaningful statistic. Managed futures funds are not an asset class but an industry. Momentum isn’t an immutable property of markets, but one that has been observable in the recent era. When an institution allocates to a momentum strategy in the hope of cushioning itself from stock market downdraughts it is really commissioning someone to sell stocks on its behalf into a falling market; no different to the failed portfolio insurance strategy that was implicated in the 1987 crash. If the odd institution wishes to protect itself in this way there is no contradiction, but if they all do, the risk of destabilising short term market behaviour will once again be high.
Winton is a leader in this industry, a beneficiary of some of the trends described and loser from others. That our industry is now an accepted part of the asset management community whilst we are a leader within it is positive. That there are now many cheap providers who implicitly question the real value offered by our services is obviously not so welcome to us. Any views we express must be understood through this prism. Winton has, however, progressively modified its strategy over recent years to be a long way from the simple trend following model embodied in a lot of alternative beta strategies. Winton explicitly regards an endogenous shock of the sort that shook the portfolio insurance industry in 1987, LTCM in 1998 or the statistical arbitrage industry in 2007 as a key risk and has taken extensive steps to mitigate it. Let’s hope all those liquid alts, smart and alternative beta, crisis offset, tail risk hedging, convexity and other portfolios and managers out there in banks and asset management companies have both the experience and the prudence to have done the same.