Fundhouse Insights Clarity Investment Blog

Limited evidence of TAA skill

February 23, 2021 by Fundhouse

As our clients know, besides reviewing fund managers qualitatively, we also go to great lengths to use data to identify skill. In today’s note, we look back at our data-driven findings with respect to Tactical Asset Allocation (TAA). Since launching in 2007, we have reviewed over 100 multi-asset investment strategies across a broad range of fund managers, amassing a significant dataset. Looking purely at TAA as a source of alpha, we find that very few asset allocation teams can evidence long-term excess returns after fees. We summarise our findings below. 

What was in scope? The range of multi-asset strategies reviewed was vast, from highly active diversified growth funds to less active balanced funds, and even included fund-of-funds and income/drawdown strategies. We also looked across risk profiles, from cautious strategies to more aggressive growth-oriented ones. We assessed each over rolling timeframes, and in all cases we shared our findings with the managers for their feedback.

How did we assess them? Each asset allocation team makes changes to portfolios based on TAA views they hold. We tested whether clients would have been better off had the managers not made these changes and simply left the portfolios as they were, or kept them aligned to a strategic allocation. Where we had the data, we also looked at how widely skilled the TAA team were (across asset classes), or whether returns were explained by a low number of views. 

What did we find?

  • Low success rate: Net of wholesale fees, TAA did not add value approximately 80% of the time. There was a fairly large group of managers that generated very modest alpha gross of fees, but costs eroded most, if not all of it. We estimate that gross of fees, the success rate would have improved from 20% to around 35%.
  • More skill in equity allocations: Many asset allocation teams appeared to be better at shifting allocations around within equities than in other asset classes. Bond asset allocation alpha was especially poor. Considering bond yields have been highly influenced by central banks, we had some sympathy with this.
  • Not enough risk is taken: More often than not, managers were not taking enough risk (credit, equity and duration) when the opportunity presented itself. There was frequently an asymmetry between how much more they valued downside protection over risk taking.
  • Alternatives let them down: Alternatives seldom provided the ‘insurance policies’ they were intended to.
  • Process questions: We often found that views based on softer concepts, like themes, geopolitics, elections, and macro variables, made it difficult to identify cause and effect between the view and the capital allocated. That is to say, correctly predicting these events is not obviously the same as correctly predicting markets.

We acknowledge that the more recent macroeconomic environment has been tough for investors, with unconventional monetary policy driving returns, and discount rates pushed to all-time lows. In recognition of this, we examined managers’ track records as far back as possible. However, we appreciate that many TAA processes were only formalised in the last 10 years (consider the recent popularity of diversified growth funds), and so the poor TAA success rates we observed were probably skewed by recent events. TAA success rates were worse between 2010 and 2020 than between 2000 and 2010, although the sample size was far larger between 2010 and 2020. 

Although the overall TAA success rate we observed has been low, some TAA processes did produce persistently good outcomes. There was an identifiable pattern to many of these approaches: they were often uncomplicated, had a focused circle of competence, were empirical in approach, and they were willing to be long-term in their outlook. They were also unafraid to add risk when their process called for it.