Tuesday, February 9, 2010

Flipping Coins Proven To Beat Professional Buy And Hold Investing

Original posted on the CXO Advisory Group web page:

Did the poor returns and high volatility of the U.S. stock market during 2000-2009 represent a tailwind for market timers? To check, we measure the performances of various simple market timing approaches (equal weighting with cash, 10-month simple moving average signals, momentum, and coin-flipping) over the decade. Using monthly closes for S&P Depository Receipts (SPY), a short-term interest rate composite and the S&P 500 index from December 1999 through December 2009 (and earlier for S&P 500 Index signal calculations), we find that:

For perspective, consider the monthly statistics for the S&P 500 Index in the following table. The mean monthly return is much lower, and the standard deviation of monthly returns higher, during the 2000s than during the prior half-century.

Next, consider a fly-off of the following simple long-only market timing strategies during the 2000s:

  1. Buy and Hold: Buy and hold SPY (as a benchmark).
  2. EW SPY-Cash: Hold equal amounts of SPY and cash, rebalancing monthly.
  3. 10-Month SMA: Hold SPY (cash) when the monthly close of the S&P 500 Index is above (below) its 10-month simple moving average (SMA).
  4. 6-1 Momentum: Hold SPY (cash) when the past 6-month return for the S&P 500 Index is positive (negative).
  5. 6-1-1 Momentum: Hold SPY (cash) when the past 6-month return, with a skip-month, for the S&P 500 Index is positive (negative).
  6. 100 Monthly Coin Flippers: Hold SPY (cash) when the coin comes up heads (tails).

For this fly-off, we make the following assumptions:

  • Use adjusted returns for SPY to incorporate dividends.
  • The return on cash is the short-term interest rate composite.
  • There are no trading frictions (biased in favor of all timing strategies). The effects of actual trading frictions depend on strategy trading frequency, bid-ask spreads, specific broker fees and account size.
  • For the 10-Month SMA and 6-1 Momentum strategies, signals derive from data just before monthly closes, allowing executions at monthly closes. Note that these strategies may be very fragile with respect to this assumption (see the blog entries of 2/3/10 and 12/18/09), such that shifting execution by small intervals may materially reduce returns.
  • Ignore any tax implications of trading.

The following chart compares the cumulative performances of the above strategies during 2000-2009. Some notable findings are:

  • 85% of Monthly Coin Flippers beat Buy and Hold, and the average Monthly Coin Flipper beat the market by 29% on a cumulative return basis. In other words, the sample period presents a tailwind for long-only market timers.
  • The best (worst) Monthly Coin Flipper outperformed (underperformed) Buy and Hold by 122% (32%).
  • The 10-Month SMA and 6-1 Momentum strategies perform similarly and beat all 100 Monthly Coin Flippers on a cumulative return basis.
  • The EW SPY-Cash strategy beats Buy and Hold but loses slightly to the average Monthly Coin Flipper.

Imposing trading frictions would depress results for all timing strategies (by differing amounts according to trading frequency). For example:

  • Monthly Coin Flipper #64 has a cumulative return 29% higher than that for Buy and Hold (same as the average Monthly Coin Flipper). Imposing a trading friction of 0.2% per change in position reduces #64's cumulative return outperformance from 29% to 13%.
  • Applying the same level of trading friction to the 6-1 Momentum strategy reduces its cumulative return outperformance from 145% to 138% (the 6-1 Momentum strategy does not trade as frequently as the typical Monthly Coin Flipper).

For a different view, we look at average (arithmetic mean) monthly returns for all tested strategies.

The next chart compares the average monthly returns for the above strategies during the 2000-2009 decade and during each year over the decade. Some notable findings are:

  • The average monthly return for Buy and Hold over the decade is approximately zero. 55% of adjusted monthly returns for SPY are positive.
  • The EW SPY-Cash, 10-Month SMA, 6-1 Momentum and 6-1-1 Momentum strategies have similarly low standard deviations of monthly returns.
  • Standard deviations of monthly returns for the Monthly Coin Flippers are lower than those for Buy and Hold.

In summary, any long-only timer of the U.S. stock market who did not beat the market during 2000-2009 has some explaining to do.

No comments: