InvestmentsNov 11 2013

Spotting bubbles before they burst

twitter-iconfacebook-iconlinkedin-iconmail-iconprint-icon
Search supported by

Historically, financial bubbles have varied in focus – from tulips to technology – but what they all have in common is the investors and their behaviour. While it is undoubtedly easier to identify bubbles in hindsight, it would be much more useful to spot them at an early stage.

The first area of contention is the predictability or otherwise of bubbles and, by extension, bursts or crashes. Statistician Nassim Taleb argues that ‘black swan’ events, which are key drivers of markets, are rare, extreme and inherently unpredictable. Another view, espoused by econophysicist Didier Sornette, is that crises are the norm rather than the exception and frequently result from bubbles, which can be modelled.

This means if we can recognise bubbles in their formative stages, it should be possible to come up with an early-warning system.

It certainly seems reasonable that in a world of many bubbles, not all will burst, with the vast majority deflating either with or without intervention. However, if the conditions that gave rise to the bubble persist or recur, the bubble could re-inflate. Bubbles may be more prevalent in early-stage markets, which lack the breadth or depth to cope with the influx of investor attention. They are not confined to particular asset classes and may manifest at stock, industry, sector or country level.

So how can we model bubbles? Mr Sornette claims investors are attracted to the exponential growth rates that characterise bubble markets and crowd into the market; herding behaviour then determines market fluctuations that increase in frequency as the bubble approaches bursting point.

His ‘financial bubble experiment’ used quantitative measures of growth and frequency, later augmented with probability, to create an index of the likelihood of future crashes. Early results were mixed with 50 per cent predictive success but the process has subsequently been refined and revised.

A more qualitative approach, the Minsky-Kindleberger taxonomy, holds that bubbles in financial markets may be divided into five emotional phases: displacement, credit creation, euphoria, financial distress and revulsion. This approach is more intuitively attractive but may be harder to pin down. The stages may be indistinct and overlap rather than being clearly defined, so deciding which phase is current is a key task.

Displacement (the initial phase) is often associated with an external shock that results in a change in opportunities for part of the economy or market. Historic examples include the spread of the internet in the US and financial liberalisation in Japan; current candidates could be the boom in emerging markets, the prevalence of social media and the development of shale gas in the US.

Credit creation, which is within the system but outside normal monetary activity, may be linked with new (and often complex) financial instruments – derivatives, such as collateralised debt obligations, for example, or liquidity injections in the form of quantitative easing.

Investors then buy in to the concept in the euphoria stage. Valuation metrics and risk premia can be revised to justify positions, critical voices are ignored and momentum trading dominates – beware the buzzword bonanza of ‘booms’, ‘paradigm shifts’ and ‘unprecedented’.

As the mania builds, leverage often comes into play. In the financial distress stage insiders start edging towards the exit, selling stock or the whole company, or borrowing at rates that seem at odds with their credit fundamentals; debt builds to the extent that it cannot be serviced by cashflows, or tax receipts in the case of governments. Unemployment, bankruptcy, defaults and forced asset sales feature as capacity is taken out. In the revulsion phase investor time horizons shorten, cashflow is paramount and perceived risk assets are shunned.

Are we seeing extremes either of valuation (bonds or defensive stocks) or of investor positioning now? Another way is to backtrack from the consequences. For instance, in asset allocation terms there may be a breakdown in the normal relationships – say, between equities and bonds – and in the bursting stages correlations (and volatility) tend to rise as assets fall. In terms of caveats, not all displacements are negative – ie, there are more white swans than black and at the cygnet stage they are all grey. We just have to avoid the ugly ducklings.

Frances Hudson is global thematic strategist at Standard Life Investments

UNDER THE BONNET: MINSKY-KINDLEBERGER TAXONOMY

Who are they?

Hyman Minsky was an American economist who specialised in the research of financial crises – what caused them, how to understand them and what can be learnt.

Charles Kindleberger was also an American economist who penned a book entitled Manias, Panics, and Crashes, about speculative stockmarket bubbles.

The five stages:

• Displacement

• Credit creation

• Euphoria

•l Financial distress

• Revulsion