Article

From rationality to realism. Book review: “A Crisis of Beliefs: Investor Psychology & Financial Fragility”

“The world…has witnessed an extensive history of financial bubbles, expanding credit, and subsequent crises…We put inaccurate beliefs at the centre of the analysis of financial fragility”.

In a new book, “A Crisis of Beliefs”, economists Nicola Gennaioli and Andrei Shleifer tackle the role of beliefs in financial markets. They highlight considerable survey evidence that contradicts the assumptions of the rational expectations (REH) and efficient markets hypotheses (EMH). Beliefs, typically an extrapolation of recent experience, are often mistaken and in predictable ways.

The authors develop an “empirically motivated and psychologically grounded model of expectation formation”. It is a step towards an economics built on “biological reality rather than rationality”. They then apply the model to account for dynamics around the Lehman crisis and to explain “credit cycles and financial fragility more generally”.

The book says little new about the psychology of expectations formation. For modelling purposes, they rely on the ‘representativeness’ heuristic developed by Kahneman and Tversky: our “tendency to judge likelihood by similarity”. This mental shortcut leads us to misjudge probabilities as we overreact to some similarity, or ‘kernel of truth’, consistent with much of the existing literature on financial bubbles.

Instead, the innovation is the use of survey evidence and the construction of a model, which, while doing away with rational expectations, maintains the rigour of micro foundations. It therefore represents an attempt to tackle head on the criticism that behavioural economics is a hodgepodge of anomalies lacking any unifying theory or analytical tractability.

These are worthy and challenging goals. The result is a credible attempt to build a better description of reality, although one cannot help but think about situations when other behavioural biases might dominate, as the authors admit. No model will ever fully capture human behaviour and adequately describe the world we live in under all circumstances; models are always simplifications, as Dani Rodrik reminds us. But some models are better than others and Gennaioli and Shleifer have taken a step down a promising avenue.

Practical applications

The book is aimed at economists or those with an interest in economic theory. Does it offer much to those with a more practical interest in financial markets?

The authors’ interpretation of the Lehman crisis and the central role of mistaken and extrapolative beliefs is entirely convincing. Quotations from leading US policymakers and evidence from Lehman’s models of house prices provide another important reminder of just how wrong our forecasts and beliefs can be.

It is a reminder to be humble and always ask ourselves what we think we really know about the future.

Equally interesting to the general reader is likely to be the presentation of considerable survey evidence, which, contrary to REH, lends support to the theory of extrapolative expectations and yields predictable forecast errors. In a nutshell, survey evidence suggests that in good times people are too optimistic and neglect downside risk; in bad times, the opposite is true.

Nothing new or Earth-shattering there you might think, but the weight of evidence is impressively broad. In assessing future stock market performance, the beliefs of individual investors, professionals and CFOs tend to be aligned and highly correlated with past returns. Such survey findings are inconsistent with a value-based approach to determining expected returns.

In the same vein, companies for which earnings expectations are highest tend to disappoint (even if this group includes a higher proportionate share of future ‘Googles’); CFOs are excessively optimistic about earnings growth when past profitability has been high; credit expansions predict crises and recessions, with a rising share of risky credit “especially worrisome”; high bank credit growth raises the risk of a crash and lowers expected returns on bank stocks etc etc.

The model of expectations the authors develop in the response to these findings has many interesting features but a couple are worth highlighting. One is the role selective memory plays in forming our beliefs. Another feature of the model is that it can generate “systematic reversals of optimism and pessimism in the absence of news”. When the news appears to stop trending, expectations can be reversed by a magnitude seemingly unwarranted by any news itself. Price movements in the absence of news are something we have long recognised and thought about, and yet media explanations of market movements are nearly always attributed to some or other ‘news’ story.

Relevance to markets today

The book presents a strong argument for treating whatever prevailing narrative exists with a high degree of scepticism. Grounding one’s expectations of future returns in a valuation framework is one way to do so – almost by definition, there is likely to be maximum hype about an asset when valuations suggest future expected returns are likely to be low and similarly the reverse is true during more pessimistic times. Survey data may complement valuation analysis by adding to our appreciation of the prevailing narrative.

What is the narrative today? This is an intriguing question and perhaps depends where you sit. In the US, equities have enjoyed one of the longest bull runs in history and economic conditions appear extremely favourable after many years of growth. We should be wary of over-optimism, you might conclude.

And yet, in the rest of the world the story seems very different, even if non-US equities have risen substantially since 2009. In Europe, sentiment feels fragile and worries over Italian politics, the integrity of the euro zone and the ramifications of Brexit dominate the headlines. In emerging markets, recurrent currency depreciations (2013, 2015 and 2018) stoke fears of previous systemic crises, while concerns over Chinese growth have remained close to the surface.

The latest surveys and commentary here, here and here would suggest widespread caution among investors rather than complacency. Moreover, in the major developed economies credit growth to the private sector has been notoriously muted following the 2008 Global Financial Crisis (GFC). If the work of Gennaioli and Shleifer is to be believed, this should provide a degree of comfort about future growth prospects in developed economies.

As we have argued before, the 2007-09 (GFC) and 2010-2012 euro sovereign debt crisis were deeply traumatic, resulting in considerable mental scarring. They remain salient in our minds many years later, and as the representative and availability biases identified in cognitive science would suggest, markets have been periodically haunted by ghosts of crises past ever since. This is a view that would be entirely consistent with “A Crisis of Beliefs”.

So, as Stuart pointed out recently, perhaps the one region where we should now guard against complacency is the US. Elsewhere, there is little evidence of complacency or over-confidence in either survey findings or the high compensation for risk implied by valuations of global equities or EM assets generally, if anything, the reverse.


The value of investments will fluctuate, which will cause prices to fall as well as rise and you may not get back the original amount you invested. Past performance is not a guide to future performance.