Market timing just does not work
By Adrian Clayton
I am not sure if you watched the SA vs Australia cricket test this summer?
Although the cricket was outstanding, it could be argued that Ricky Ponting’s luck in winning the toss was at times the most significant difference between the teams. In fact, Ponting won all six tosses. Every one.
I don’t have to tell you that a 100% toss win rate is simply unheard of in cricket and although coins might land on their side, normally one would have only a 50% chance of being right when flipping a coin. In fact, the probability of the 6 out of 6 is only 1.6%. Putting it differently, had another six tosses occurred, the chances of many of these going against Ricky would have escalated.
This brings me to my point.
It is during volatile times such as these in which we presently find ourselves that many investors begin to dispel tried and tested investing practices and, instead, are attracted to strategies that are erroneous, rely on luck but are intuitively attractive.
One such approach is market timing. Market timing occurs where managers claim an ability to perfectly time which asset classes will outperform others and try to aggressively switch in and out of these asset classes at will and allege a perfect or high success rate. The objective of such a strategy would be to offer investors maximum gains when stock markets are rallying and no downside when they are falling. A further element of this approach is that such a manager’s claim to be able to do so over short time periods – that is periods of less than one year. As one can imagine, with 386 out of 470 domestic unit trusts in South Africa (excluding the cash, bond and Fixed Income Varied Specialist funds) having produced negative returns over the past 12 months and the average return being down about 16%, the attaction of an ‘all weather, bullet proof fund’ is presently in vogue – in a big way.
At Alphen we believe that aggressive market timing is a failed strategy and we base this argument on four factors.
Academic Research. What is great about our industry is that data is readily available. Take for example the unit trust/mutual fund industry in America. The first mutual fund was the Massachusetts Fund launched in 1924, a very long time ago. With so much and such lengthy data streams, it is possible to intensely scrutinize successful versus unsuccessful strategies over long time periods. In 1986, Brinson, Hood, and Beebower (BHB) published a study about asset allocation of 91 large pension funds which they analyzed for the period 1974 to 1983. They replaced the pension funds’ stock, bond, and cash selections with corresponding market indexes and found that the indexed quarterly returns were higher than pension plan’s actual quarterly returns. In other words, these large pension fund managers added limited value by asset allocating; it would have been better for these funds to have stayed in a set or passive asset blend over time.
Whilst this research is compelling, at Alphen we believe that marginal asset tilts around a set asset benchmark can add value, but someone who undertakes binary shifts, or “jumps” from the one asset to the next, may well end-up with worse results than maintaining a consistent passive plan. Space and time do not allow for further discussion of other academic papers on market timing, suffice to say that we have yet to find any academic work that demonstrates that anyone anywhere has shown any ability to add value over long periods of time through aggressive market timing.
Assessing what great investors have done through the ages. Whilst all great investors have bought shares and other asset classes over time and also sold them when they deemed this appropriate, their approach was invariably based on valuation metrics and less so on short-term market sentiment phenomenon. The likes of Benjamin Graham and David Dodd’s skepticism for forecasting were so pronounced that they even shied away from predicting future company profitability. Much of their work involved backward looking profit analysis of companies and determining whether their current-year profits were realistic within the context of their long-term profitability. They affectively normalized company earnings.
Professor Robert Shiller has received great acclaim for his work on this too. My colleague, Neels van Schaik, has written on this topic many times in past Alphen Angles, if you seek further insights on normalizing company profits when analyzing a company. The point we make here is that the greatest investors of all time dispute anyone’s ability to predict future events with any degree of clarity, that they have based much of their investment philosophy and process on historical work.
On this note, I thought some quotes from a few of the world’s investing gurus would clarify this point.
Gensler and Baer in “The Great Mutual Fund Trap”
John C. Bogle
Our own experiences. I have worked in markets for fifteen years. In the bigger scheme of things this is a relatively short period, but I have been fortunate to work with some brilliant minds. During this development I been privy to enormous successes achieved by colleagues but, unfortunately, in certain instances outright unparalleled success has sometimes been followed by shattering failure. In each instance when success has been sustainable, colleagues have relied on a set investment approach based on long-term fundamental investing and their success has relied on clients willing to ‘play ball’ and give them time. On the other hand, I have yet to work with anyone who has successfully implemented a timing approach with any degree of success, over a long enough time period, to make it statistically relevant. In fact, what I can vouch for, is that the “timers” I have encountered are prone to unbelievable highs and accompanying industry acclaim, followed by severe underperformance and accompanying embarrassment. With respect to my own experimental attempts to buy and sell shares during my first few years in this industry, I would at best claim a 50% hit rate, not good odds when other people’s monies are at stake!
Finally, this thesis of timing being a failed concept is based on Common Sense. This brings me back to my flipping the coin concept. If we know based on academic research, empirical evidence, tried and tested guru strategies and our own experience, that timing offers at best a 50/50 chance of being successful and that one should not rely on luck as an investment strategy, then surely it makes sense to seek a strategy where the odds are in one’s favour?
So what is the correct approach to asset allocation?
At Alphen we believe that the wisest asset allocation approach is not too dissimilar to the proven BHB methodology. Every asset class that one can invest in has a risk which is relatively measurable. By combining assets in intelligent ways, it is possible to construct solutions where risk is relatively quantifiable and it is even possible to offer some insight as to return characteristics. Such solutions should define the long-term objectives of the product or portfolio and quantify the time the manager needs to achieve this objective and thus how long the investor should remain invested.
Many of these solutions can be created catering for differing risk and return needs or characteristics. Instead of trying to manage one fund, or product or portfolio that aims to be everything to everyone (in other words attempting to be an equity fund when markets are rallying and a cash fund when markets are falling), the fund should rather operate within a set risk band where its return characteristics are more predictable and can thus be aligned with an investor’s specific needs.
This is a credible method of money management and aligns the task of the advisor to that of the asset manager and prevents the creation of deluded expectations.
Source: Adrian Clayton, Alphen Asset Management, March 27, 2009.
Performance Optimization WordPress Plugins by W3 EDGE