Skip to content

Global Tactical Cross-Asset Allocation: Applying Value and Momentum Across Asset Classes: summary and key takeaways

The paper ranks 12 asset classes each month on momentum and a simple yield-based value measure, then compares top and bottom quartiles. Momentum and value each earned positive long/short returns, and a combined score earned 11.9% a year, with results holding after several adjustments.

Blitz, D.C.; Vliet, P.V. · 2008 · The Journal of Portfolio ManagementSummary updated 20 November 2026
Published2008Journal article
Asset focusMulti-assetTAA
TypeBothTactical Allocation
Reading time5 minthis summary

Global Tactical Cross-Asset Allocation: Applying Value and Momentum Across Asset Classes at a glance

The question
Can the value and momentum patterns known within asset classes also be used to rank whole asset classes against each other for tactical allocation?
Dataset & period
12 asset classes (US large-cap, US mid-cap, US REITs, UK, Japan and emerging markets equity, US Treasuries, US investment-grade, US high-yield, German and Japanese government bonds, US cash). Monthly data, February 1986 to September 2007. Pre-sample test on 8 asset classes, January 1974 to January 1986.
Method
At each month end the 12 assets are ranked on 1-month return, 12-1 month return and a valuation score. Valuation is the earnings yield for equities and the yield-to-maturity for bonds, each net of the local risk-free rate, with fixed adjustments for structural biases. Assets are split into four quartile portfolios of three, and the long top and short bottom quartile (Q1-Q4) portfolio is held for the next month. The combined score weights 1-month momentum, 12-1 month momentum and valuation at 25%, 25% and 50%.

Key findings of Global Tactical Cross-Asset Allocation: Applying Value and Momentum Across Asset Classes

  • The spread between the top and bottom quartile (Q1 minus Q4) came to 6.9% when sorting on 1-month momentum, 7.9% on 12-1 month momentum and 7.9% on valuation, each significant at the 1% level.
  • The combined strategy returned 11.9% a year on the long/short portfolio, with volatility of 10% and an information ratio of 1.2.
  • The two momentum strategies correlate at 0.3 with each other and negatively with the valuation strategy, which supports combining them.
  • The 12-1 month cross-asset premium is close in size to the US stock momentum factor (UMD, 8.4% over the same sample). The cross-asset value premium is larger than the US stock value factor (HML, 3.4%).
  • After adjusting for each portfolio's average exposure to the 12 asset classes, the combined Q1-Q4 return stayed at 11%.
  • The 12-1 month momentum alpha loses significance after controlling for the stock momentum factor, while the combined strategy alpha stays at 11%-12% a year across CAPM, three-factor and four-factor adjustments.

Robustness, as reported in the paper

  • Transaction costs of 10, 25 and 50 bps per trade were tested. The 1-month momentum strategy loses about half its performance at 10 bps, while the combined strategy keeps 5% a year at 50 bps.
  • A pre-sample test on 8 asset classes for January 1974 to January 1986 gave somewhat lower but still strong returns.
  • Volatility-adjusted positions (scaled to a 10% target using 60-month volatility) gave the same information ratio of 1.2 for the combined strategy.
  • Results were not driven by one asset class, and alpha spreads were positive across regimes based on VIX, term spread, credit spread and interest rate level.

Caveats and limits

  • Transaction costs are excluded from the headline results, and futures were not actually available for every asset in the sample.
  • The approach cannot easily include asset-specific variables, so it favors breadth over depth, as the authors note.
  • The valuation adjustments are fixed and were chosen to remove structural biases; the authors say assets should not all be valued by these simple measures in equilibrium.
  • The authors say chance or data mining and time-varying risk premiums cannot be fully ruled out, and the sample ends in September 2007 (inferred).

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

Why it matters

The paper shows a single-model way to compare very different asset classes using only returns and simple yields. It also gives turnover and cost sensitivity by signal, which helps judge how implementable each leg is. The authors suggest it could serve as a standalone or an added building block in a tactical allocation process.

Questions about this paper

How does the paper measure value across different asset classes?

It uses earnings yield for equities and yield-to-maturity for bonds, net of the local risk-free rate, with fixed adjustments so no asset class is structurally favored.

Does the combined strategy survive transaction costs?

The paper reports that the combined strategy keeps 5% a year even at 50 bps per trade, while the 1-month momentum leg is most sensitive to costs.

Is the cross-asset momentum effect just the known stock momentum effect?

The 12-1 month version is largely explained by the US stock momentum factor. The combined strategy alpha remains at 11%-12% a year after the factor adjustments.

Where can I read the original paper?

The original is at its publisher: https://doi.org/10.3905/jpm.2008.35.1.23. 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.