Data Diary

Measures of risk appetite – it’s a lovely view from up here

Posted on 21 March 2010

Risk appetite is the most fickle aspect of the valuation process.  From exultant fist-pumping on the mountain top to indecisive cowering in a dark cave, market sentiment can swing swiftly or, just as likely, can defy seemingly rational expectations for nigh on an eternity. What is the value in tracking risk appetite then?

1) Extremes in risk appetite often signal the exhaustion of a trend

Our market-based measures of risk appetite are coincident indicators, that is, changes in investor sentiment will be immediately reflected in prices and/or volumes. When aggregate investor sentiment is leaning heavily one way or t’other, the relevant measure of risk appetite will reflect this.

Particularly in the shorter term then, extremes in sentiment often signal that the current trend is vulnerable. If the market is overweight a consensus view, the risk of a stop-fuelled counter-trend move is heightened.  For this reason, indicators such as the CBOE put call ratio are often useful as short term windvanes – consider the current market position:

Once again, the ratio is testing lows that generally signal that bullishness is at an extreme and caution is warranted.

Volume can also be useful in this regard.  Consider this chart of the S&P500 and 10dayMA trading volume.

Spikes in trading volume higher generally occur at or around price lows in the index.  Notably, the common criticism that the current rally has been built on declining volume (and therefore is not confirmed by volume) is pretty clear from the above.  Just that when placed in this longer term context, it looks like rallies are more likely to be lower volume that selloffs regardless of their place in the economic cycle.

2) The absolute level of risk spreads can indicate which sectors are likely to outperform

To state the bleeding obvious – the absolute level of risk spreads are a reflection of the prevailing environment – the challenge is to decipher whether they are appropriate for the future environment.  Changes in risk spreads are after all key drivers of prices.

For example, in a credit bubble world, where interest rates are low, cash is plentiful and ‘benign’ is the prevailing mood, risk spreads of all hues will be bid in as money chases returns. The risk curve flattens, credit spreads are low and stable while equities outperform.  Conversely, when the risk curve is steepening due to rising defaults and with financing costs on the rise, equities will underperform. An interesting study in this regard by EconoPic illustrated that the absolute level of risk spreads might be a good indicator of 1 year forward returns as between equities, high yield debt and investment grade debt.

So what we want is a do-it-yourself roadmap, so we can assess which sectors are likely to outperform if risk spreads move to reflect that world. In this context, the current level of risk spreads are expressing some optimism. Credit spreads (as between AAA and investment grade corporates) are threatening levels not seen since the last days of the bubble, while the VIX has returned to the lows of May 2008. These are suggesting that we are well and truly through the crisis.

If you believe that that we are tracking to muddle through with global liquidity being maintained to support the recovery, perhaps you can argue that credit spreads and the VIX can keep tightening. For mine, in a world where the recovery has peaked, and the threats from sovereign debt defaults, rising trade tensions and ongoing assets problems are clear and present dangers, the ongoing push up the risk curve looks less likely to be favourably rewarded from here on in.

In short, I’d see the fair value risk curve as being north from here – and steeper.  For example, this means a range trade on the AAA-Baa spread of between 100 and 150 bps.  Pushing further out the risk spectrum, equities are also likely to be defining the top end of a range – with a material risk of a deeper risk aversion trade that takes us back to the long run moving average of the S&P’s P/E multiple.