The axiom is that most invested $'s can't beat the market gross, from which it would follow fairly readily that fully publicizing a method which will certainly beat the market by a significant margin would soon lead to it no longer beating the market, since investors would gravitate to it, and the axiom would insure its demise as a successful strategy. IOW the axiom can fairly easily be extended to 'easy to understand and implement strategies which will significantly and certainly beat the market cannot be widely known'. But there's no axiom by which they can't exist, nor that they can't be known if there's uncertainty they'll continue to work, which is virtually always the case for any strategy which has or is claimed to have beaten the market.neurosphere wrote:I haven't read the whole thread, so forgive me if this has been answered. But can you show me which mechanical system has been shown to outperform the market?Runalong wrote:On a personal level, sound quantitative analysis can (and has and does) effectively outperform if tied to a mechanical system that excludes emotions, such as "hold for exactly one year (or x # of months) no matter what then sell (unless it still qualifies on that date)".
An example of a simple mechanical strategy which beat buying and holding the S&P in at least one fairly prolonged period:
http://www.cboe.com/micro/buywrite/Pap- ... eb2012.pdf
Selling a 1 month at-the-money (ATM) put option on a given notional unit of the S&P (say for example a put on 1 ES futures contract instead of being long 1 ES contract, which is in turn approximately equal to being long 500 shares of the SPY ETF) each month, holding the rest of the notional amount of investment in t-bills*, gave a total return of 10.8% with standard deviation of return of 10.2% in period end 1988- end 2011, v 9.1% total return with std deviation of return of 15.0% of the S&P itself, per this study. That's pre-tax, so applicable to tax advantaged accounts, but the option strategy has less favorable tax treatment than index buy/hold in a taxable account.
I'm not actually sure this strategy should even count as 'non Boglehead': it's based on a whole index, does not involve market timing or stock picking, is pretty simple for investors to do themselves with basic options knowledge and execution skill and has relatively low transactions costs, though it requires at least monthly attention.
There's also a plausible answer for this strategy to the question the basic axiom always poses: who is going to lose out relative to the index if said strategy is to win out? And the plausible answer here is that active trading institutions (and perhaps individuals) are too eager to buy puts to cut their exposure particularly in times of turmoil, leaving a volatility risk premium, which this research suggests extends up to ATM strike (it's pretty obvious for significantly below-the-money strikes), which the strategy then harvests. But note that this strategy has strictly less risk than buy and hold index. Both the index holder and ATM put seller lose by the decline of the index in the month, but the put seller's loss is cushioned by the premium received, likewise limited to the month's premium on the upside. So it's expected the std deviation of return would be lower than holding the index. The apparent market anomaly is that the put sell strategy can match (let alone slightly exceed) the index return.
However, like virtually any such strategy, the fact that it worked from 1988-2011 doesn't prove it will work from 2012-2035. The previous success of the strategy could militate against its future success, or conditions could just change. Come to think of it, that's not a lot different than 'small value tilting' in that respect. That doesn't seem to catch much flak around here even from the hardcore, but is arguably also a mechanical system claimed to have often beaten the (whole market) index.
*an individual would get still higher returns, by up to several 10's of bps in today's market, by holding some of the cash in highest paying FDIC insured money market account rather than t-bills, in today's market paying up to 0.95% v a few bps yield on a 1 month t-bill, for basically the same risk.