Skip to content

Lazy Momentum with Growth-Trend Timing: Resilient Asset Allocation (RAA): summary and key takeaways

The paper builds Resilient Asset Allocation (RAA) from the earlier Lethargic Asset Allocation (LAA) in three steps: a five-asset equal-weight portfolio, a faster market filter and a slower unemployment filter. Over the full backtest RAA shows a higher CAGR and a smaller maximum drawdown than LAA and than a 60/40 benchmark, with turnover that stays low.

Keller, W.J. · 2020 · SSRN Electronic JournalSummary updated 20 November 2026
Published2020Working paper
Asset focusMulti-assetTAA
TypeBothTactical Allocation
Reading time4 minthis summary

Lazy Momentum with Growth-Trend Timing: Resilient Asset Allocation (RAA) at a glance

The question
Can a more diversified near-static portfolio, timed by unemployment and a fast crash-protection signal, beat LAA and 60/40 on return and drawdown while trading rarely?
Dataset & period
US-listed ETF assets (QQQ, IWN, IEF, TLT, GLD, plus canaries VWO and BND), with proxy data for the early years, monthly rebalancing, full sample Dec 1970 to Nov 2020. Benchmark is a static 60/40 mix of SPY and IEF. Sub-periods: in-sample, out-of-sample, last 20 years, last 10 years.
Method
The strategy holds five assets at equal weight (QQQ, IWN, IEF, TLT, GLD). It moves to a 50/50 mix of IEF and TLT only when two signals are bearish at once. The first signal is US unemployment above its level 12 months earlier. The second is a fast momentum filter on the canary assets VWO and BND, which turns bearish if either asset turns bad. Rebalancing is monthly.

Key findings of Lazy Momentum with Growth-Trend Timing: Resilient Asset Allocation (RAA)

  • Full-sample CAGR for RAA is 12.3%, against 9.7% for the 60/40 benchmark.
  • Maximum monthly drawdown falls from 29.5% for 60/40 to 11.8% for RAA.
  • Against LAA, full-sample CAGR rises from 11.5% to 12.3% and maximum drawdown falls from 15.0% to 11.8%.
  • The Ulcer Performance Index improves from 1.73 for LAA to 2.83 for RAA over the full sample.
  • Average cash fraction rises from 12.2% for LAA to 15.5% for RAA, and yearly transaction cost rises from 0.03% to 0.16%.
  • RAA does worse than LAA in the last 10 years on return (8.5% vs 9.7%), but its drawdown is smaller there (5.8% vs 8.4%).

Robustness, as reported in the paper

  • Results are shown for in-sample, out-of-sample, 20-year and 10-year windows as well as the full sample; RAA beats LAA on return in all of them except the last 10 years.
  • Transaction costs are included, based on a one-sided fee of 0.1%.
  • The author reports that the equal-weight five-asset static portfolio has a smaller drawdown than the four-asset LAA portfolio even without timing.

Caveats and limits

  • Several parameters (breadth setting, unemployment filter, cash assets) were chosen on the in-sample period, so the out-of-sample window is only partly independent (inferred from the paper's description).
  • Early-period data for the equity ETFs rely on proxies, and the author calls pre-1970 data quality weaker.
  • The backtest rebalances monthly to equal weights; the author only guesses that less frequent rebalancing would make little difference.
  • Results are a historical simulation on one set of US assets, with no live trading record (inferred).

Items marked (inferred) are our reading of the design, not statements by the authors.

Why it matters

The paper shows how a slow macro filter and a fast market filter can be combined so that a strategy stays invested most of the time. It gives a transparent rule set that practitioners can compare with other tactical allocation designs on return, drawdown and turnover.

Questions about this paper

When does RAA move out of the risky portfolio?

Only when US unemployment is higher than 12 months earlier and at least one canary asset (VWO or BND) shows a bearish fast momentum signal. It then holds IEF and TLT at 50% each.

How does RAA compare with 60/40 in the paper?

Over the full sample, RAA has a 12.3% CAGR against 9.7% for 60/40, and a maximum drawdown of 11.8% against 29.5%.

How much does RAA trade?

The paper puts average trading at about one month per year. Yearly transaction cost is 0.16% over the full sample, higher than the 0.03% of LAA.

Where can I read the original paper?

The original is at its publisher: https://doi.org/10.2139/ssrn.3752294. We link to the paper and do not host the PDF.

Follow new quant research

Download MacroFinance market data as CSV or Excel, or pull it by API, and test the ideas in these papers yourself. New summaries are added as we review them.

Create a free account
Summaries are written by MacroFinance from the published papers, in our own words, for information only. They are not investment advice and may omit detail. Papers are not hosted here: each page links to the original at its publisher.