Week 51 Day 2: Timing the Market: Missing the Best Days

JPMorgan's 'Guide to Retirement' data (2024): $10,000 invested in the S&P 500 from January 2003 to December 2022 grew to $64,844 if you stayed fully invested. If you missed the 10 best days: $29,708 (less than half). If you missed the 20 best days: $17,826. If you missed the 30 best days: $11,701 (barely better than cash). The 10 best days: 7 of them occurred within 2 weeks of the 10 worst days. March 24, 2020 (best day in years) came 9 trading days after March 12, 2020 (one of the worst). You cannot capture the recovery if you fled during the crash. The lesson: stay invested. Always.

Market timing fails for two reasons: (1) you have to be right twice -- when to get out AND when to get back in. Getting out feels easy during a crash (panic sells itself). Getting back in is nearly impossible because the recovery begins when things still look terrible. In March 2009, the market bottomed while unemployment was still rising, banks were still failing, and every headline screamed doom. By the time things 'felt safe' (late 2010-2011), the S&P 500 had already recovered 80%. Investors who waited for safety missed most of the recovery. (2) You incur real costs every time you sell: capital gains taxes, transaction costs, and the psychological barrier to re-entry (once you sell, admitting you were wrong and buying back in is emotionally agonizing). Dalbar's research shows the average equity fund investor earned 4.3% per year over 30 years while the S&P 500 returned 10.7%. The 6.4% gap is almost entirely explained by market timing -- buying after rallies (greed) and selling after declines (fear). The mathematical proof: if you could perfectly time the market (selling before every decline and buying before every rally), you would earn approximately 20-25% per year. But studies of actual market timing attempts -- including newsletters, technical analysis systems, and professional tactical allocation funds -- show net returns that are consistently lower than buy-and-hold. Hulbert Financial Digest tracked market timing newsletters for 30 years and found that none -- literally zero of the hundreds tracked -- consistently beat a buy-and-hold strategy after costs.

The academic evidence against market timing is extensive and unambiguous. Sharpe (1975) showed that a market timer must be right approximately 74% of the time just to match a buy-and-hold strategy, due to the asymmetric return distribution of stocks (a few very large positive days drive most of long-term returns). Jeffrey (1984) refined this estimate to 69-80% depending on the frequency of timing decisions and transaction costs. Likelihood of achieving this accuracy: virtually zero. Neely, Rapach, Tu, and Zhou (2014) conducted the most comprehensive test of market timing using 14 predictor variables (dividend yield, earnings yield, interest rates, term spreads, etc.) over 1927-2012. Their finding: combining all 14 predictors into an optimal forecasting model generated a statistically significant but economically trivial timing benefit of 0.5-1.0% per year before transaction costs and taxes, which was fully consumed by trading costs in practical implementation. Individual variables performed even worse -- most had negative out-of-sample forecasting ability. Kinnel (2014) at Morningstar documented the 'behavior gap' for every fund category: in every asset class studied (U.S. stock, international stock, bond, allocation), the dollar-weighted return (what investors actually earned) was lower than the time-weighted return (what the fund earned) by 0.5-2.5% per year, reflecting the systematic tendency of investors to buy after gains and sell after losses. The behavior gap is the single largest cost in the average investor's financial life -- larger than fees, larger than taxes, larger than any single investment mistake. Eliminating it requires one thing: stop trying to time the market.