When you are interested in taking a long position in a security, is it better to wait for the price to fall from its highs (a “pullback”) or to take the position immediately at current prices? I decided to analyze this question using a performance forecast for SPY under different circumstances.

I created a simulation that forecasts the performance of SPY at different time horizons and different times of initiation of the position. I did one run in which the position is initiated immediately, and another in which the position is initiated only after the price has fallen a certain percentage from the highs (a “pullback”). The simulations all take an initial position worth $100, and the position size varies according to the entry price. So the entry price in the “no-pullback” simulation is always $100, which equates in the simulation to exactly 1 share of SPY. The “pullback” entry prices can be more or less than $100, and so the quantity purchased varies depending on the price (such that EntryPrice * Quantity = $100). The results are displayed below.

The main takeaway from the simulation is that, broadly speaking, waiting for a pullback in SPY does not improve returns even though the position size is larger. In the case of the 2% pullback, the difference seems to attenuate with a holding period of 15 years or more. For all the larger pullbacks I tested, the difference never goes away even with a 50 year holding period. Part of the reason for the difference is the pullback-entry position has less time in the market to rise in value, and part of the reason is that waiting for a pullback causes one to systematically avoid taking a position in the best possible circumstances. In other words, the best performing markets are the ones that never have a pullback, and thus you would never enters since you would be waiting to enter on a pullback. You see this in the simulation results: In the simulation with the 20% pullback entry strategy, the first few results are all zero, indicating that in those simulation runs a 20% pullback was rare or nonexistent, and thus time in the market was missed to such an extent that entire holding periods were lost.

Based on this work, it seems to me that if you expect a position to increase in value, you ought to take the position without waiting for the price to fall first. It seems that the value of time in the market tends to outweigh the value of accumulating a larger position by waiting for prices to fall. If the position neither gains nor loses value, then I think waiting for a pullback is irrelevant; you would expect the results to be the same either way. Finally, if you expected the position to fall in value, you would lose less money waiting for a pullback simply because you were out of the market during the initial price fall.