Some of the tactical asset allocation strategies that we track have significant exposure to rising interest rates, or more specifically, to the types of assets that are most negatively affected by rising rates. While we don’t (yet) track every published TAA model, the strategies that we do track are broadly representative of the TAA space, so I think it’s fair to draw some broader conclusions about TAA in general based on our data.
To try to capture TAA’s exposure to rising rates, below I’ve shown the 28 strategies that we track, as well as each strategy’s average allocation to the five most vulnerable asset classes those strategies trade. Not all exposure is the same, so I’ve colored green those assets the strategy is designed to avoid when that asset underperforms, and I’ve colored red those assets that the strategy cannot avoid (green is preferable to red). I’ve also included an aggregate score (low numbers are preferable to high). More on both the color coding and aggregate score in a moment.
|Average Allocation to Assets Most Sensitive to Rising US Rates
Use the Arrows to Sort This Table
|Adaptive Asset Allocation||7%||29%||0.43|
|Composite Dual Momentum||13%||9%||0.35|
|Davis’ Three Way Model||33%||0.66|
|Elastic AA – Defensive||47%||0.94|
|Elastic AA – Offensive||41%||0.82|
|Faber’s GTAA 13||4%||6%||4%||1%||0.19|
|Faber’s GTAA 5||15%||0.15|
|Faber’s GTAA Agg. 3||9%||5%||3%||1%||0.27|
|Faber’s GTAA Agg. 6||8%||6%||6%||2%||0.30|
|Faber’s Ivy Portfolio||20%||0.40|
|Faber’s Sector RS||0.00|
|Flexible Asset Allocation||17%||0.17|
|Glenn’s Paired Switching||47%||0.94|
|Protective Asset Alloc. CPR||3%||35%||1%||0.42|
|Protective Asset Allocation||3%||52%||1%||1.11|
|Robust AA – Aggressive||7%||0.07|
|Robust AA – Balanced||15%||0.15|
|US Equal Risk Contrib.||35%||31%||1.32|
|US Max Diversification||58%||4%||1.24|
|US Max Sharpe||25%||25%||0.50|
|US Min Correlation||33%||24%||1.14|
|Varadi’s Min Correlation||30%||1.20|
|Varadi’s Perc. Channels||40%||0.40|
More about this table:
Color Coding: Green is preferable to red.
Consider two hypothetical strategies. The first buys TLT (long-term US Treasuries) when TLT closes above its 12-month average. In that case, TLT would be colored green, because the strategy is designed to avoid TLT when it’s underperforming and should have less exposure during an extended period of rising rates.
The second strategy buys SPY (S&P 500) when SPY closes above its 12-month average, otherwise it buys TLT. In that case, TLT would be colored red, because whether or not TLT is underperforming has no bearing on the trade, and the strategy may have just as much exposure during a period of rising rates.
Aggregate Score: Low numbers are preferable to high.
In an attempt to summarize this data, I’ve included a very “back of the envelope” aggregate score for each strategy. The aggregate score is the sum of the following for all assets: (average_allocation * asset_sensitivity * avoidance_score), where asset_sensitivity equals 2 for TLT and 1 for all other assets, and avoidance_score equals 2 for red cells and 1 for green.
This is an imperfect measure for sure, but it does a reasonably good job of identifying extremes. Use the up/down arrows in the table above to sort by the aggregate score.
Most of the strategies with the highest exposure are of the more static variety: the All-Weather Portfolio, Varadi’s Min Correlation, and three of the portfolio optimization strategies (MC, MD, and ERC). All lean heavily on bonds without a way to rotate away when bonds underperform. In the case of Min Corr and the portfolio opt strategies, allocation is based on volatility and correlation (ignoring returns), meaning all will continue to signal a similar allocation, even if bonds turn sour.
Protective Asset Allocation stands out as a very dynamic (and historically high performing) strategy that also has a high aggregate score. That’s because it frequently uses intermediate-term US Treasuries (IEF) as a “fall back” asset, regardless of IEF’s performance. In response to this, over the next week we’ll be adding an alternative to PAA from Keller and Keuning’s original paper that has the ability to rotate between IEF and cash. Even though we haven’t officially launched this strategy yet, I’ve included it on the table as “Protective Asset Alloc. CPR”. Note the improved aggregate score.
Strategies with the lowest aggregate score tend to have the ability to rotate away from bonds when they underperform: a number of Meb Faber’s strategies, and the Robust Asset Allocation and Flexible Asset Allocation strategies all stand out.
We provide a lot of ways for members to assess tactical asset allocation strategies, and in the coming year plan to continue doing an even better job of that with more advanced analytical features. But certain characteristics of strategies, like exposure to an extended period of rising rates, are difficult to capture statistically (*). I hope that these ad hoc figures shed a bit more light.
We invite you to become a member for less than $1 a day to track all of these tactical asset allocation strategies in near real-time, or take our platform for a test drive with a free limited membership. Have questions? Learn more about what we do, check out our FAQs or contact us.
(*) …regressing portfolio returns over historical interest rates aside. We’ve opted to take a less sophisticated, but I think more useful approach in this post. Many of the strategies that we track do not have history that extends to the market’s last extended period of rising rates, meaning an apples to apples comparison is difficult.
A note on cash and short-duration bonds:
Some of the strategies that we track were originally designed to include exposure to short-duration bonds like SHY. When replicating those strategies however, we always replace that short-duration bond exposure with “cash” (equal to the 3-month UST rate). Even though an instrument like SHY may result in superior performance historically, it’s not really applicable to today’s market. Given the frequency with which most of these strategies trade, rates are too low for these instruments to generate enough return to overcome transaction costs.
Had we tracked each of these strategies exactly as originally designed, exposure to instruments like SHY would be another factor to include in the table above. But because we’ve replaced that exposure with cash (which reduced historical performance) the future effect of rising rates on the cash asset will actually be positive. Not hugely positive mind you, because most of the strategies spend little time in cash, but certainly not negative.