Simple and Effective Market Timing with Tactical Asset Allocation Part 2 - Choices
12 Pages Posted: 5 Mar 2018 Last revised: 27 Apr 2018
Date Written: April 18, 2018
In Part 1 a simple market timing algorithm was described that switches from an exchange traded fund representing U. S. equities (SPY) to one holding treasury long bonds (TLT) every month on the last day, the switch being made to whichever ETF has the greatest ratio of current adjusted closing price to adjusted closing price µ months earlier. The parameter µ was determined so as to maximize total return and minimize the total number of trades. It was demonstrated that, over the 10-year period ending on 12/31/13, this model produced a cumulative annualized growth rate (CAGR) and a maximum monthly (daily) draw down, respectively, of 13.9 and 17.3 (19.0)%; note that these values differ slightly from those in the original paper because a more conservative estimate of trading costs is now included.
In the present work, this model is continued to the present day but also modified by extending the two-ETF comparison to one that includes 5 ETFs. The 3 new ETFs considered track MSCI developed market equities outside the U. S. and Canada (EFA), the NASDAQ 100 (QQQ), and MSCI emerging markets (EEM). The extended model is shown, to date, to deliver 74% higher CAGR than the two-ETF model, but with no increase in draw down despite the increased volatility of the 3 new ETFs in comparison to the two original.
A third option is also examined in which an additional constraint is added to the algorithm, namely that unless all 5 ETFs produce positive returns over the -month period, the end-of-month choice reverts to cash or a cash surrogate. If the cash surrogate chosen is the ETF tracking the 7-10 year treasury bond (IEF), the maximum monthly draw down to date is shown to be less than half that of the two-ETF model and the CAGR to be actually slightly better.
Keywords: market timing, algorithm, model, tactical asset allocation
JEL Classification: C51, C52, C53, C58, G11
Suggested Citation: Suggested Citation