Investors and risk managers are staring into the abyss. Unable to use past asset pricing models to predict the immediate future they are, says Professor Riccardo Rebonato of EDHEC Business School, hamstrung by a toxic combination of politics, the current global crisis and asymmetry of risk. This, he says, is the time they should look for ‘the best-looking horse in the soap factory’ to help them at least stay in the race.
Prices matter. When the prices of goods and services become untethered from supply and demand, the results are long queues and shelves full of products nobody wants to buy. It is no exaggeration to say that the advantage that price information gave Western markets substantially contributed to the demise of the controlled-economy of the then Soviet Union. Prices, in the end, may have contributed to winning the Cold War more than cruise missiles.
Prices of financial instruments also matter. In a well-functioning market, they convey information about the expected growth of the economy and about risk premia. When markets are efficient, intelligent investing boils down to exploiting in the most efficient manner the risk compensation afforded by exposure to the priced risk factors. Without price information, this task becomes impossible.
Unfortunately, the prices of financial assets are currently no more closely linked to economic fundamentals than the prices of the unwanted tins of food were on the Soviet shelves of the 1970s. This state of affairs has been caused by the monetary actions of central banks the world over. Perhaps central bankers had no alternative, and certainly I would not have known how better to handle the 2008 financial debacle, and the current Covid crisis. The fact, however, remains that the price distortions caused by these interventions are unprecedented – and their consequences extremely far-reaching.
All assets in the firing line
As the 2000s drew to a close, and rates approached (or went through) the zero bound, central bankers began engaging more and more widely in the so-called ‘unconventional monetary measures’ – in other words, quantitative easing (QE). As the spectrum of financial instruments that have been purchased by central banks has widened, in maturity and in the asset-class dimension, a ripple – or substitution – effect has propagated from one type of asset to the next. Even assets that are not directly bought by central banks feel and adjust to the yield compression of the direct targets for purchase. The prices of virtually all assets have therefore been affected. How long will these distortions endure?
The original strategy behind QE could be described in fewer than 10 words as ‘the creation of controlled asset bubbles’. The idea (or the hope) was that, as the economy recovered, these bubbles could be gracefully deflated – rather than suddenly bursting. The deflation, of course, was to be achieved via a steady and gradual rise in rates back to ‘normal’ levels: gradual enough for the economy not to take fright, but steady enough to ultimately eliminate the distortions.
Unfortunately, a mixture of asymmetry in risk (bad surprises are far more frequent than unexpected cheques in the mail), political pressure from overbearing politicians, and plain Covid bad luck has made the progressive QE interventions act as a ratchet, with no significant unwinding after each new bout of purchases. As a result, prices of virtually all financial assets currently bear no relationship to their economic linchpins, the expectation of future growth and the compensation for risk.
This has devastating consequences for investors. The wedge between real-life expectations and the price of a given asset is no longer a rational compensation for risk, but a reflection of how active central banks have been in purchasing that asset, or its close substitutes. Risk premia are everywhere close to non-existent or negative. At the time of writing, the level of the S&P500 equity index is very close to where it was immediately before we learnt that a few people in China had become infected with a novel coronavirus. This price level cannot be plausibly explained in terms of unchanged expectations of future cash flows, or unchanged risk premia. It can only be understood through the lens of the QE-induced price distortions. Reading information about expected economic growth or risk premia has become a thankless, and close-to-impossible, task.
Finding an investment North Star
Investors, of course, can still hope to make money by investing now, in the hope that future prices will become even more distorted in the future – in the words of investor and fund manager Bill Gross, they can still engage in the risky game of finding ‘the best-looking horse in the soap factory’. There is no denying, however, that the traditional compass of fundamental value analysis is these days being subjected to a QE-induced magnetic storm of such magnitude as to make it unusable.
The challenges for risk management are no less severe. As we are in uncharted territory, past history can provide very little guidance as to how this experiment will end. Statistical techniques, of which value at risk (VaR) is the best, but not the only, known example, are based on the implicit assumption that the future looks, in some complex way, like the past. If we have no past to look back upon, how can these frequentist methods help us?
When it comes to ‘tail risk management’, usually I am a strong proponent of what I call ‘coherent stress testing’. This involves using our understanding of how the world works to obtain conditional estimates of future outcomes: if such and such happened, then it is likely that this and that will unfold. However, this approach relies on our economic theories, enabling us to assess what is likely to happen and what is not. Unfortunately, asset pricing models based on the traditional notions of expectations and risk premia currently offer close to no guidance.
Will the price distortions become so severe that the full edifice will eventually buckle? And, if this happens, which last straw will break the back of the QE camel? I don’t even know how to begin to answer these questions. In conditions of deep uncertainty, fall-back decisional criteria (such as minimax) tend to come to the fore. These are blunt and crude decisional instruments. Alas, currently they too may be the best-looking horses in the risk management soap factory.
