Works but not worth it

Volatility-targeting plain SPY, VTI and VUG: close, not enough

· 3 min read

The idea

Scale an equal-weight mix of SPY, VTI and VUG to a 10% volatility target, in the spirit of the volatility-managed portfolios in Moreira and Muir (Journal of Finance, 2017). No trend filter. Their paper scales by the inverse of last month's variance rather than a fixed 10% target; the 10% level is ours.

What happened

Sharpe Yearly return Worst drop SPY buy and hold
2017–2021 0.89 11.5% 18.3% Sharpe 0.92, 33.7% drop
2022–2023 0.35 6.9% 10.7% Sharpe 0.12, 22.1% drop

The pass rule allowed a Sharpe shortfall of at most 0.02 against SPY. In 2017–2021 it was 0.035.

The official runs

These are the official default-cost results for the pre-registered vol-targeted core funds rule, one of the 15 trials in the literature batch. Sharpe and yearly return credit idle cash at the 3-month T-bill rate, like the figures already on this page. Worst drop is from the simulated book as traded. The SPY rows are reference points, not trials; the same-days row is frictionless, which slightly favours it. Its own rule was judged against fully invested SPY, not against the 200-day overlay.

2017–2021SharpeYearly returnWorst dropTime invested
Vol-targeted core funds Pre-registered · Failed its own rule0.8911.54%18.3%77%
SPY, same invested days Reference0.8911.26%17.1%77%
SPY, fully invested Reference0.9218.28%33.7%100%
2022–2023SharpeYearly returnWorst dropTime invested
Vol-targeted core funds Pre-registered · Passed its own rule0.356.87%10.7%55%
SPY, same invested days Reference0.296.07%9.1%55%
SPY, fully invested Reference0.124.26%22.1%100%
Growth of one dollar for vol-targeted core funds, in 2017–2021 and 2022–2023.
Growth of $1. Vol-targeted core credits idle cash at the 3-month T-bill rate, the same way as the yearly returns above. The SPY mixes are frictionless. Official windows only; the January 2022 embargo between them is not shown.
Drawdown from peak for vol-targeted core funds, in 2017–2021 and 2022–2023.
Drawdown from peak. Vol-targeted core uses the simulated book as traded. Official windows only.

Against fully invested SPY, the book's Sharpe was 0.035 lower in 2017–2021 (the bar allowed a shortfall of 0.02) and 0.22 higher in 2022–2023. Its worst drop was 54% and 48% of SPY's in the two windows. The drawdown chart shows where that cushion came from, and the growth chart shows what it cost in 2017–2021.

Why it failed

A near miss against a bar we set in advance is still a miss. The pass rule allowed a Sharpe shortfall of at most 0.02 against SPY; in 2017–2021 the shortfall was 0.035 (0.89 against SPY's 0.92). It did better in 2022–2023 (0.35 against SPY's 0.12), but one passing window is not enough when both had to pass. Across the seven strategies in that batch, the best holdout deflated Sharpe was 0.40 (bar 0.95), so even the better window doesn't carry much weight once we count the tries.

Receipts: the rule was pre-registered before the run.