Five years ago, (March 2014) Daniel Kahneman filled the 2,300 seats of the Methodist Hall in London at prices averaging well over £20, marketing his book Thinking Fast and Slow – a book about biases in human rationality. The expensive seats were crowded out with hedge fund managers and their broking hosts. Bizarre as this may seem, there was no escaping the popularity of behavioral finance.
Right now, hedge fund strategies based on inefficiencies generated by human trading decisions are in an unhappy state: the rolling debate on the death of momentum has been rudely suppressed only by an even more urgent debate on the death of value. Should we care? Absolutely. This is not just a knife fight in a phone box about hedge funds: these questions are driving problems across the whole liquid investment universe. If trading decisions now sit mostly with computers, particularly ones which run routine index matching algorithms, then Kahneman’s web of human biases will play a much smaller part in the market pricing, and this, in turn, will leave less investment opportunity for our brilliant pointy heads who supposedly float above these biases. The next iteration is for all this genius to profit from the errors of its own ways, but though there is a certain amount of that going on, it’s a much tougher gig. If, as an investor, you want the benefits of this type of inefficiency be warned that entry to the club is mostly by invitation only, and the price of membership is both truly eye watering and rising.
Much as this narrative about computers may appeal for its proximity to so many similar stories about the role of technology in modern society, it could still be gloriously wrong about the death of value and momentum. Note, for example, that problems with both these factors started to emerge soon after the 2008 crisis. Monetary policy changed radically then (while computers did not), and as the pivotal driver of financial markets, policy may be no less a suspect than computers.
For momentum, the stable and relatively homogeneous global rate structures of the last ten years have reduced the number, length and strength of trends. This can hardly be surprising, given that the risk-free rate is the foundation for the pricing of most financial instruments. Trend following strategies can take larger positions in independent moves than they can in highly correlated ones, and they do particularly well in the longest and strongest moves so a FED that explicitly targets asset price stability (at least downwards) thus wilfully confounds the best opportunities and the number of independent moves.
There is, of course, the argument that crowding is a bigger problem. If so, then perhaps one would expect to see trends retrace further when they turn: too many trend followers would extend the trend but then also magnify the reaction as they all try to get out again. But evidence of this is scarce. It is therefore perhaps more natural to conclude that it is on balance, more the shortage of mice than the plethora of mousetraps that is the problem. In short, it looks as if post-2008 monetary policy has been bad for momentum.
As for value, well, it is often suspected that this works better with higher levels of bond yields, that is, with a modicum of inflation. It is also thought to prefer positive sloping yield curves (not an especially demanding caveat since positive curves are so common as to be called normal). But in recent years, curves have been flattening as another consequence of monetary policy which includes bond yield ‘repression’. In Japan, a place where value has been in a hole for seemingly ever, the BOJ even targets this flatter curve. But Japan is just an extreme case of a global phenomenon and the cumulative underperformance of value since 2009 now exceeds that of the great Tech Bubble at the turn of the century.
If computers were the sole root of the issue, then one might expect all such equity factors to suffer equally. But the low beta (or low vol) factor has been performing positively. Perhaps it is as much a beneficiary of the policy regime (as investors look for bond substitutes in the most stable equities) as value and momentum are victims.
Supposing then that at least some portion of the difficulties with these two key factors, value and momentum, may indeed be attributable to monetary policy rather than computers, the ‘what next?’ question becomes more interesting, for while computers are indisputably here to stay, those who have an ‘end of history’ type view of the monetary policy regime are on altogether thinner ice.
Of course Central Banks like to keep things nice and stable. But the tectonic plates beneath them are always shifting. As the pressures build for seismic change, Central Bank responses become less intuitive, less robust and more extreme until in the end there is a quake and everything falls over for a while. Things were ever thus. The managed exchange rate regimes of the ERM and Asia as recently as the 1990s were a very formal instance of this, with their periodic busts, most notably in 1997. This grand process is cyclical.
We know that the growth sub cycle now is just turning up again: leading indicators are recovering, while a series of Chinese stimulus measures have yet to bite. Meanwhile, US unemployment claims are at 50 year lows and with Trump looking for his second term, fiscal policy is unlikely to be restrained. Right now, inflation is dipping, but oil is higher and it is much less clear how this story all plays out later in the year. So we are approaching a crunch point, the like of which we have not seen in the last decade: something will have to give. Is this really a good time to bet that it is the computer and not monetary policy at the heart of our value and momentum drought?
Hedge fund performance in April was muted in the face of rising equities and lower volatility in other asset classes. The best performing strategy was managed futures, with longer-biased managers in equity long-short and credit also benefitting from the benign market environment. Relative value strategies were mixed, in particular strategies driven by the value factor struggled as markets were led higher by growth names.
Managed futures performance was positive again following a noteworthy March, but with returns in April driven by their long equity exposure as opposed to the dominance of their long bonds exposure in March. The month started with more of a mixed profile for managed futures managers as bonds reversed some of their March move in the first half of the month, but as bonds stabilized generally manager performance improved. By the end of the month, managers were generally still longer more risk in bonds than equities, despite the performance of equities during April, and also significantly long the dollar. Commodity exposure remains fairly light.
In equity long-short, there was a mix of returns, but with winners during the month significantly outperforming losers this meant that the average return for the strategy was positive. The best performing managers were either longer-biased (benefiting from the positive return in equities) or sector focused (who seemed to be able to capitalize on the strong risk-on move better than their generalist peers). Small/mid cap specialists also performed positively in general. The worst performing managers were the broad market generalists with some form of value bias. Equity managers continue to note the difficulty of trading in a low-volatility, low-rate environment, and feel that the return of more accommodative policy from central banks has supported the share price of otherwise poor quality or ‘expensive’ securities, making shorting harder and market efficiency worse.
This phenomenon spilled over into the quantitative equity market neutral space, where risk premia and fundamental strategies built from factor models both performed poorly during the month (albeit with a high degree of dispersion between different managers). Shorter term and more technical strategies in statistical arbitrage generally held up and finished the month with small positive returns, as did the suite of ‘new finance’ quantitative strategies such as machine learning.
Credit managers saw a continuation of the Q1 environment into April, with positive returns across the board for investment grade, high yield and leveraged loans, despite the slight pull back in treasuries during the month. In particular, leveraged loans were a source of notable returns versus traditional credits, and within credit lower quality outperformed higher quality, another sign of a risk-on month. Under such a backdrop, the directionally exposed managers in distressed debt outperformed, mainly due to market returns but also due to positive resolutions of well-held distressed companies. Preferreds also had a positive month, particularly financial preferreds, off the back of a notable reporting season for financials.
Credit long-short managers also benefitted from the risk-on move, as managers generally held a bias towards being long the credits with wider spreads. Structured credit was also positive as a quiet market backdrop meant that managers generally posted carry-driven returns for the month.
April was a relatively quiet month in event arbitrage. Merger deal flow continues to moderate and deal spreads remain tight. The ongoing trade negotiations between the US and China along with uncertainty around Brexit are weighing on the corporate M&A activity. That being said, risk arbitrage managers are still finding adequate opportunities to deploy risk. One notable M&A development in April was the competitive bidding situation between two oil industry companies for a third. The merger spread between two of the companies responded around the news of a counter bid and thus created a more attractive opportunity for managers. In special situations, managers remain active as corporations are increasingly looking for corporate structure enhancements to support their growth trajectories.
Discretionary macro managers had a mixed month for April, with those managers who were long risk-assets clearly outperforming their peers. Other winners were those with a structural long dollar exposure, both from a developed markets or emerging markets perspective. Despite the positive moves in equities and the USD, most managers continue to bemoan the lack of volatility across all asset classes. In particular the return of bond volatility to lower levels is hampering returns from fixed income relative value specialists.