PsyFi Search

Wednesday 3 February 2010

Buy and Hold, the Least Worst Option?

Trigger Happy Traders?

The arguments over market timing versus buy and hold are likely to continue until the end of time. Often these debates aren’t really over the actual merits of one approach or the other but centre on the philosophy of the methods.

Proponents of timing strategies can’t understand why anyone would ever adopt buy and hold as it’s “obviously” a worse method of investing. Meanwhile supporters of buy and hold regard market timers as a bunch of hopeless goons with itchy trigger fingers. As ever, the truth is sitting about one inch behind your eyes.

Marketing Timing Topers

There are two arguments against market timing. The first, roughly, is that it’s impossible since there are no discernable trends that can be used to accurately time the market. The best that can be done is to roughly identify cheap stuff and then hold it for long periods waiting for commonsense to assert itself.

The second is that even if market timing works investor psychology makes it virtually impossible to implement successfully for the majority of people. Here the philosophy is more vague but the evidence less so – most investors throw away any excess gains they might make in an unstructured and hopeless set of random trades. The problem isn’t that investors can’t beat the market, it’s that they can’t beat themselves.

Short-Term Shirt-Losers

Most market timing arguments revolve around short-term trading. As I’ve argued before, in Rise of the Machines, trying to beat the securities industry at a game that’s vastly weighted in their favour is pointless. In any case, most people don’t have the time to spend seeking out short-term opportunities because they’re too busy doing economically productive things like raising children, surfing the internet or actually working for a living.

However, there are smarter and longer term timing strategies that do seem to work reasonably well. A combination of factors like CAP/E, Tobin’s q factor and momentum effects have a decent record in this regard. Personally I’m sceptical that such methods could survive a concerted attempt by the investment industry to take advantage of them, suspecting that they might disappear like other anomalies once they became popular. Still, the current evidence supports the fact that they work and ignoring this because it’s an uncomfortable fact getting in the way of an otherwise beautiful theory would be irrational.

Contracyclical Value Chasers

Indeed there’s decent evidence for the success of all sorts of contracyclical investing strategies based on various value effects. The problem with all of these is not that they don’t work but that they can go wrong for nastily long periods while other rather more growth oriented stocks go off on a bender of quite extreme proportions. Both of these tendencies then trigger behavioural effects on people that lead – ahem – to less than optimal investing approaches.

So the argument for a buy and hold strategy doesn’t rest on it being better than value-based timing strategies. No, it merely revolves around the unfortunate reality that it usually provides better long-term returns than the market timing strategies that people actually use. It’s not that it can be guaranteed to provide especially good returns compared to other stock investing approaches but that it’s the least worst option once you take most people’s psychology into account.

Improving Investors, Constraining Advisors

The standard responses to this position are either that this requires investors to receive better training or that the marketing power of the securities industry needs to be restrained in order to reduce its impact on investor decisions. The evidence suggests that neither of these methods is likely to work.

Most financial training doesn’t seem to have any effect and when it does do something it usually goes on to produce exactly the opposite effect to that desired. Meanwhile restricting the investment industry’s ability to market itself is highly unlikely to work. Quite aside from the intense political lobbying that such a suggestion would generate the evidence from the tobacco industry is that every attempt to reduce marketing power has exactly the opposite effect: reminding people of the diseases that cigarettes can cause actually triggers nicotine cravings in addicts, rather than putting them off the idea.

The Evidence for the Prosecution

Turning to the actual evidence for investors’ market timing addictions we find a fascinating analysis of mutual fund cashflows by Friesen and Sapp. The idea behind the study is to examine when investors put money in and take it out of mutual funds in order to analyse whether these timing effects improve or retard their investment performance. Looking at over 7,000 mutual funds over a 13 year timescale they find:
“investors underperform by about 0.13% per month, or 1.56% annually, relative to a buy-and-hold strategy"
However, this gets better (or worse, depending on your perspective). Much, much, much better.

And The First Shall Come Last

Among the curious effects that Friesen and Sapp tease out of their data is the highly remarkable one that the worst market timing performance is associated with the most successful funds. That’s worth repeating – the best performing funds are correlated with the least successful investor performance. This is a stunning, even by the standards of behavioural finance, example of snatching inglorious defeat from the jaws of seemingly unstoppable victory.

Trying to figure out what’s going on here leads us into a labyrinth of ill-founded speculation. However, the obvious conclusion is that these investors are so addicted to market timing that they happen upon the best performing funds purely by luck and trade away their potential gains on the same basis.

Momentum and Value Effects

This isn’t by any means the only interesting finding. They also show that timing underperformance is positively correlated with momentum style funds but is negatively correlated with value type funds. So investors in momentum funds tend to (unsuccessfully) market time while those in value funds don’t.

Again we can only speculate but as momentum style funds tend to follow stocks that have shown recent momentum, and are therefore vulnerable to mean regression effects, it’s possible that fly-by-night investors seeking the latest hot trend will be more attracted to them. Value funds seemingly are of more interest to buy and hold type investors aiming to capitalise on the other side of the mean reversion equation, suggesting that they are at least vaguely aware of the concept of “buy low, sell high”.

Passive or Active?

There’s an immediately obvious connection to the different behaviours of passive index style investors and active fund adherents. The former are presumably buy-and-hold merchants while the latter active traders?

Well, no, as it happens. The researchers saw the same trading effects regardless of the type of fund investors used. It looks very much like activist investors are quite happy to use index funds to implement trading strategies.

Just not very well.

Behaviourally Biased

The lesser effect of trading on value funds is also mirrored on bond and money market funds. It seems that it’s recently outperforming equities, with their rollercoaster emotional highs and lows, that are the main locus of these trading effects. There’s also a strong suggestion that less sophisticated investors are more likely to suffer from these problems as the research offers a casual side-dig at advisors:
“... our evidence suggests that those investors who are most likely relying on advice from a broker perform especially poorly from a timing standpoint.”
So the overall finding is that return chasing investors only manage to catch underperformance, typically investing while markets are high and selling while they’re low and thus managing to catch the troughs between the waves with predictable consequences. It’s hard to explain these results other than through the medium of behavioural psychology; another brick in the wall we’re trying to build. It’s beginning to look formidably high.


Related Articles: Don't Overpay For Growth, It's OK To Lose Money, Investing In The Rear View Mirror

14 comments:

  1. I never heard of CAP/E Could you please explain what it is?

    ReplyDelete
  2. It seems to me that its impossible for the masses to beat the market (through timing, etc.) for the same reason that its impossible for the NFL to win more than half its games.

    For every team with a winning record, there will by definition be a team with a losing record.

    Suppose that all investors engaged in market timing. For every investor who timed successfully, another would time unsuccessfully. Add in transaction costs, and this would be a negative-sum game for investors.

    ReplyDelete
  3. James - CAP/E is the Cyclically Adjusted Price Earnings ratio - essentially a P/E adjusted for variations over a period of years - 5 or 10 is the common period. The idea is that it smoothes out peaks and troughs of over and under valuation. Probably Robert Shiller is the leading current proponent although it was (I believe) originally proposed by Ben Graham.

    Parker - agreed, it's logically impossible for a majority of people to successfully time a market with guaranteed counterparties. However, as about 75% of men think they're above average drivers we shouldn't expect that to stop people trying. Perhaps the more interesting question is whether anyone can consistently time the market.

    ReplyDelete
  4. Personally I’m sceptical that such methods could survive a concerted attempt by the investment industry to take advantage of them, suspecting that they might disappear like other anomalies once they became popular.

    I agree that the edge that long-term timers now possess would disappear if this strategy were widely promoted. However, I see this as a feature, not a bug.

    If large number of investors were selling when prices got too high and buying when prices got too low, we would achieve price stability. Stock prices would always go up by about 6.5 percent real per year (the gain that is justified by the economic realities). We would not have insane bull markets nor the insane bear markets that inevitably follow from them.

    Could anything be better? We would continue to enjoy that juicy 6.5 real long-term average return without having to experience the biggest risk of stock investing -- seeing your portfolio wiped out in a crash. We also would avoid the economic crises that are caused by the wealth devastation we see in bear markets.

    You did a great job with this one, Tim. It is extremely hard to write in a balanced way on this topic.

    Rob

    ReplyDelete
  5. Very well-balanced article...Great job!It is not that easy to write on this topic unbiasedly and you did full justice to it. Hope to read more...keep writing on such issues.

    ReplyDelete
  6. As James, I also never heard of CAP/E. Can you please tell us what it is?

    ReplyDelete
  7. I agree with Rob. I too think that promoting this strategy widely would swipe the edge possessed by long-timers but it is nothing that cannot be mended. It is not a bug that could destroy the entire system...

    ReplyDelete
  8. Thank you Timarr for explaining CAP/E. I had a doubt about it. It's clear now.

    ReplyDelete
  9. On market timing:

    I think its quite possible to tell when the market is cheap or expensive, compared to its historical valuations.
    However, it seems to me that the usefulness of this is limited by 2 factors:

    1) Historical averages may not hold in the future. There is no iron law that sets the equity premium to be as high as it was in the 1900's, for instance.

    2) Even if historical averages hold true, then there may be an opportunity cost in waiting for reversion to the mean, which could take decades.

    ReplyDelete
  10. An asset allocation strategy would reduce the volatility of a portfolio. Rebalancing to cash when equities had outperformed and back in when they had fallen.
    It wouldn't eliminate the problem but would mitigate the damage of big crashes.
    You just have to have faith to buy a falling market, and then wait.

    ReplyDelete
  11. Rob Bennett -

    That doesn't make sense.

    The fact that investors are irrational is exactly the reason we can get a good return on equities above non-risky assets, and that so much of the world's wealth stays in non-risky assets.

    If all participants were purely rational, the equity risk premium must logically disappear as any gain above non-risky assets would be bid down by the arbitrage your hypothesis depends on.

    And that is what roughly what happens, right? As the equity risk premium appears to disappear in a bull market (stocks can only go up!), more investors get pulled in, and bubbles appear. As the market tries to correct those new investors pull out (buy high! sell low!) and the correction is oversold.

    Lemondy

    ReplyDelete
  12. Fabulous article, snappily summarized by Parker John.

    You need to work out the utility cost/benefit of the fact that I *like* trading away some of my gains in the active part of my portfolio, because of the thrills!

    Perhaps that's the missing chunk of the equation. Give me a credit on your Phd. ;)

    ReplyDelete
  13. I am new to this forum, but I am optimistic and enthusiastic about your subject mater, as I am a vocal proponent of behavioral economics.

    I would suggest you take a look at some of Mebane Faber's work on market timing. His paper entitled "A Quantitative Approach to Tactical Asset Allocation" offers a rigorous and compelling argument for a simple multi-asset timing strategy. Of course, I agree that alpha deteriorates over time, but the nature of timing strategies are such that big money generally can not take advantage of it. This gives the small investor an advantage that he or she should attempt to capitalize on.

    I would appreciate your feedback on this paper in the context of your argument.

    ReplyDelete
  14. Hi Gestalt

    Thanks for drawing this to my attention. All papers require some concentrated attention to try and extract the essence of them, and I'll try and do this at some point, but a quick skim suggests that what Faber has here is a empirically backed approach without a theory behind it.

    Now such an approach requires that we try and understand why it's successful - there are too many examples of science ignoring the actual evidence to blythely dismiss any evidence based results. However, there's also the reality that many anomalies disappear as soon as investors recognise they exist.

    To be blunt, as a matter of dogma I doubt there are any magic bullets for investors. However, I do agree that small investors can navigate their ways between the gaps if they're nimble enough. That I'm not that skillful is no reason to disparage those that do: I'll take a look at this in more detail as soon as I get some time.

    ReplyDelete