Risk and roulette

19 June 2014

Risk is often always viewed as undesirable. But without risk, active fund management can never generate the necessary superior performance to justify its existence: no traditional long only manager could beat the market; no absolute return manager could beat cash. Diversifying away all risk is therefore never appropriate for an active fund manager: without risk, we are condemned to perpetual mediocrity.

To the outside world, the stock market is often seen as incorporating the kind of binary risk seen on a roulette wheel, yet every active fund manager through the superiority of their investment philosophy and process should (mistakenly or otherwise) believe that the odds of superior performance are stacked in their favour. As such, the analogy of the professional gambler managing risk on a biased roulette wheel can illuminate the subject of portfolio risk management.

If every active fund manager needs to take risk, management of risk should be about isolating and diversifying away only inappropriate risk.  This means that we must define what constitutes desirable or appropriate risk. Since every active fund manager should have an investment philosophy which explains, in their opinion, what moves share prices and therefore what they always look for as desirable characteristics in stocks, diversification away from this style bias is also never appropriate: the value manager should always invest in cheapness, the growth manager in quality and the earnings surprise manager in stocks with superior earnings surprise potential. As such, it is always appropriate that volatility emanating from a fund manager’s style bias is diversified at the fund selection level rather than through an individual manager’s style drift.

We should also note that style drift is most likely to occur when the manager is unable to attain immediate reward for their investors. In order to avoid style drift it is better that the manager lowers initial expectations. Although we believe (although others may not) that our investment religion - earnings surprise - is a superior and more consistent investment philosophy than its obvious peers, value and growth, and this is what both distinguishes us from and diversifies our returns from our peers, this does not mean we expect to be permanently rewarded: in fact, we accept that through taking risk, it is not possible for any active fund manager, unless engaging in a fraud, to never suffer periods of underperformance or drawdowns; though clearly we would expect these to be more than offset by times when our funds are delivering out-performance and attractive positive returns. We have high conviction in our ability to generate performance for our unit-holders, but this is very different from a belief that we will do this all of the time. Even when their style of management is out of favour, the professional investor is - somewhat ironically - managing risk appropriately by always adhering to it: in their opinion, it should always give the best chance, though not a guarantee, of future outperformance.

This is better explained in an analogy. Imagine that a professional gambler has exclusive access to a casino in which the only game is roulette, where the ball spins on a biased wheel, where instead of there being an equal number of red and black numbers, there were three times as many red numbers than black.1 Betting on red would always maximise - though never guarantee - the gamblers chances of being correct. If such a roulette wheel existed, the gambler would, over time, earn a significant reputation as being able to consistently beat the house. He would be feted by the gambling press as being a “star” gambler, even though his activity could not be described as gambling in any conventional sense.

Yet suppose the last three times the croupier span the wheel the ball dropped on black. There are likely to be murmurings that the gambler had lost his edge and that maybe it was time to bet on black. There may even be accusations, particularly in the Financial Times, that the gambler was simply “lucky” previously and that it was never possible for anyone to beat the house. Others might offer well-meaning advice that the gambler should reduce the size of his stakes until the ball starts dropping on red again. Unless the gambler was very weak minded he would realise that his chances of being correct on the next spin of the wheel were still three times higher in still calling red and would continue to call red.

Appropriate “risk management” of the biased roulette wheel isn’t changing the call from red to black, but ensuring that the stakes on each spin of the wheel are always conservative enough to ensure that capital can survive a streak of unlikely adverse outcomes, and also ensuring that conviction in bets was consistent through the size of his stakes, so that he did not indulge in a “gamblers fallacy” of believing that the odds of the next spin of the wheel could be influenced by recent results. Risk management for the gambler never involves taking zero risk – even his biased roulette wheel loses a quarter of the time – but ensuring that betting on a roulette wheel loaded in the gamblers’ favour isn’t turned into a loss making activity through loss of a disciplined and rational approach in response to near term adversity.

It is however always appropriate to diversify away inappropriate risk and as such the biggest undesirable risk in our portfolios - given our style bias - is always the failure to correctly identify earnings revision trends, either through faulty analysis of individual companies or of wider macro-economic factors. As a result our biggest concern is always whether our views on particular stocks are legitimate. Given human imperfection it is impossible that we will ever perfectly execute our process: to construct a concentrated but diversified portfolio of companies that will over the near future subsequently uniformly see superior earnings surprise. We will also make mistakes believing in potential in stocks that transpires not to exist. As such it is always appropriate to diversify our investment risk across industries, countries and stocks, but only to the extent that superior earnings surprise can still be identified in all of our investments.

Diversification for its own sake is never desirable. By contrast, conviction in earnings surprise potential in an individual stock in a sector or industry where the catalysts are stock specific is always more valuable. However, although we will invest in any industry we can understand and identify potential, it is a fact of life that earnings surprise potential is unlikely to be equally spread at all times across industries, countries and stocks. As such indirect biases to sectors and countries are always a risk that can be managed, without ever being fully diversified away.

We also believe that running concentrated portfolios is desirable risk. The identification of earnings surprise is not easy: it involves significant labour intensive research and often contrary thinking. We want to be able to focus our research efforts on our “best ideas”. When we find these opportunities we want these stocks to have a meaningful impact on the portfolios and not be diluted by inferior opportunities. It is sometimes argued that high conviction fund managers who only invest in stocks they actually believe will make money for their unit-holders are too risky, given the necessary deviation and volatility from market index weightings entailed. On the other hand it is relatively uncontroversial to suggest that equity index positions, if considered desirable, can be more cheaply bought - though after fees not replicated - through tracker funds and ETFs. Value for money in the fund management industry should never involve paying active management fees for even partly passive products. A diversified portfolio of skilled high conviction funds will likely both manage risk and deliver superior performance after any fees.

Some of our funds have different objectives: this is important in assessing what constitutes appropriate and inappropriate risk. Most obviously, our traditional long-only Alpha fund will nearly always be fully invested with an aim of beating the market, whilst our long/short Absolute Return product has an explicit mandate to manage volatility and will therefore always have a more modest market exposure. In order to maintain a consistent level of risk of the Absolute Return fund, we will increase our market risk in periods of lower market volatility and decrease our market risk when market volatility is high. The aim of our traditional long-only funds in beating the market requires simply consistently high execution of our earnings surprise investment process. Alpha will most probably outperform Absolute Return over a stock-market cycle but at the cost of higher volatility.

It’s also time for the fund management industry to come clean: risk means inconsistency of return. But far from being a uniquely undesirable characteristic, risk – where taken appropriately in accordance with a coherent and proven investment process - is in fact the raison d’etre of the active fund management industry. Just as the professional gambler with the biased roulette wheel manages the risk of unlikely but adverse outcomes by maintaining a consistent and disciplined approach, consistent exposure to desirable risk is also the single biggest factor behind avoiding undesirable portfolio investment risk.

Barry Norris
Argonaut Capital
June 2014

1 A typical roulette wheel will have 18 black spaces and 18 red spaces. However, a European wheel will also have a single zero and an American wheel a double zero in addition. As a consequence the bet in red or black is not quite binary (48.6% on a European wheel and 47.3% on an American wheel). This illustrates that without a bias to the wheel (in our analogy a proven investment process) the long-term returns from roulette will be negative (in our analogy akin to a tracker fund taking no risk after fees)

This website is for UK professional investors only. It is not intended for any person who would qualify as a Retail investor. Past performance is no guarantee nor a reliable indicator of future results and the content on this web page is a marketing communication from Argonaut Capital Partners LLP which is not intended to be viewed as a piece of independent investment research.

The material on this web page and the wider Argonaut website is not intended to be substitute to the full legal documentation for any product or products mentioned and to any information which investors must obtain from their financial intermediaries acting in relation to their investment in the products mentioned on this website.

The views expressed on this website are those of Argonaut Capital Partners and do not constitute any form of advice and the information on this website does not form any part of any contract for the sale or purchase of any investment.

Information contained in this website is thought to be accurate at the time of publication however is subject to change. Argonaut is under no obligation to update you of any future changes. The unauthorised access, copying or re-use of the information contained in this presentation by any other person is strictly forbidden.

Please see Argonaut’s Terms & Conditions prior to reading the information on this web page.