Four Important Details in Tactical Asset Allocation

Without leverage, our ability to profit on the trade is severely diluted.

A similar effect occurs with cross-asset trades. Even for two asset classes that share similar risk levels, accounting for unintended risk-factor exposures can be important. Consider going long a high-yield bond index and short a 5-year U.S. Treasury index with the intention of capturing a declining credit spread. Without explicit duration matching of the indices, there can be residual interest rate exposure. This latent exposure can be particularly harmful if we consider that rates and spreads can exhibit significant negative correlations during negative economic shocks or flight-to-safety periods.

Factors and style premia applied within an asset class can, largely, ignore these effects as risk factors are generally shared and stable. The same is not true for asset classes, and blindly applying the same approaches without acknowledging this difference can lead to unexpected results.

Tactical Often Increases Internal Portfolio Concentration

While cross-asset dynamics can play an important role in the construction of tactical portfolios, they can also play an important role in the question of, “should we bother being tactical at all?”

This was the topic of a commentary: Rising Correlations and Tactical Asset Allocation[5].

In the commentary, we used a simple example of a static 50/50 stock/bond portfolio, and a tactical strategy that can flexibly allocate between stocks and bonds. Our question was simple: when is it better to hold the tactical strategy and when is it better to just hold the static portfolio?

In many ways, the answer is a function of available diversification. Note that the TAA strategy will always be more concentrated in one asset class than the static benchmark is. Or, conversely, the TAA strategy will always be less diversified. 

When correlations are high between stocks and bonds, there is little diversification benefit foregone by the increased concentration in the TAA strategy. On the other hand, when diversification opportunities abound, the hurdle rate for TAA to add value above-and-beyond a well-diversified portfolio increases dramatically.[6]

Note that this is not necessarily true for factor investing at the stock level. Switching from a passive equity index to a long-only value portfolio can actually introduce beneficial diversification benefits, as the equity beta exposure remains, but an “active beta” (i.e. the value strategy) is added as well.  We saw this effect in our recent commentary Factors & Financial Planning.[7]

As somewhat of a tangent, an interesting byproduct of the changing hurdle rate for TAA is to “time our timing,” i.e. dial the magnitude of tactical decisions within the portfolio based upon internal diversification available within the benchmark policy portfolio declines. In Improving on risk parity[8],researchers from J.P. Morgan used a similar concept for dynamically allocating between a risk parity and mean-variance optimization process. They found that when there was significant dispersion between asset class Sharpe ratios, the forecast risk required by MVO was worth bearing, while it was better to use risk parity when Sharpe ratios converged.

Diversification is an important hedge against forecast uncertainty. TAA explicitly foregoes diversification in the pursuit of return. The trade-off, however, is not always straightforward. The time-varying nature of asset class dynamics means that an identical trade could have dramatically different hurdle rates for success depending on when it is made.

There may be times that diversification is so abundant that doing nothing is the best course of action. 


As a firm that specializes in systematic tactical allocation, we believe strongly that active approaches can lead to more efficient portfolio constructions, particularly when based upon established style premia such as value, momentum, carry, and trend.

That said, we also recognize that utilizing style premia in a multi-asset fashion can introduce complexities, that when unaddressed, can lead to unexpected (read: poor) performance.  Caveat emptor: understanding how a manager addresses these problems is critical for establishing long-term expectations.

Corey Hoffstein is the Co-founder & CIO at Newfound Research, a participant in the ETF Strategist Channel.

[1] For the sake of brevity, we’re not going to comment on the ongoing efficacy of price-to-book, whether price-to-book is really applicable across all sectors, and whether there are structural sector-based considerations that need to be made when measuring “cheapness” or “richness.”

[2] For convenience, we’re just assuming that stocks within each leg are equally-weighted.

[3] The one exception here would be the low-volatility factor, but firms like AQR apply leverage to the lower volatility leg for this exact reason.

[4] We’re playing a bit fast-and-loose here with duration assumptions.


[6] This ignores the reality that the tactical signals employed may themselves be more accurate when correlations are lower; e.g. relative momentum tends to favor bonds during market crises, which is also when we tend to see negative correlations emerge.