Strategic & Tactical Asset Allocation
Tactical Tilt Impact Calculator
Enter your policy portfolio weights, then specify a tactical tilt (overweight or underweight per asset class). The calculator computes the expected return impact of the tilt, the resulting tracking error against the policy benchmark, and the information ratio if the tilt view proves correct. All calculations run in your browser.
Direct Answer
A tactical tilt impact calculator measures the expected return impact, tracking error, and information ratio of overweighting or underweighting specific asset classes relative to a policy portfolio benchmark. Enter your policy weights and a tactical tilt to see how much added return the tilt would need to justify the tracking error it introduces if the view proves correct. Results are illustrative only, not investment advice.
How to use this tool
Start with your strategic policy weights, the long-run target allocation from your Investment Policy Statement. Then enter a tactical tilt: positive numbers overweight an asset class vs. policy, negative numbers underweight it. Tilts must sum to zero (the total portfolio is still 100%).
The tracking error is the annualised standard deviation of the difference in returns between the tilted portfolio and the policy benchmark. A tilt of +5% equities / −5% bonds produces tracking error approximately equal to σ(equity − bond) × 0.05 = √(σ_eq² + σ_bond² − 2ρ_eq,bond × σ_eq × σ_bond) × 0.05. Higher active deviations and lower cross-asset correlation both produce higher tracking error.
The information ratio tells you how much expected return per unit of tracking error the tilt provides, conditional on the tilt view being correct. Most institutional TAA programmes target an information ratio above 0.5 for approved tactical tilts. An IR below 0.3 suggests the expected return improvement is not large enough relative to the active risk taken. See Tactical Asset Allocation Signals for how to build the return view that drives this calculation.
Frequently Asked Questions
Why must the tilts sum to zero?
A tilt reallocates within a fully invested portfolio, so anything added to one sleeve has to come out of another. Tilts summing to a positive number would describe a portfolio holding more than 100% of its capital, which is a leverage decision rather than a tilt. Requiring the sum to be zero keeps the comparison clean: the tilted portfolio and the policy portfolio hold the same total capital, so any difference in outcome comes from the reallocation alone.
What does the tracking error figure measure here?
It is the annual standard deviation of the difference between the tilted portfolio's return and the policy portfolio's return, computed from the volatility and correlation assumptions entered. It describes how far the tilted portfolio is expected to diverge from policy in either direction, not how much it might lose. A tilt with high tracking error can still be a small absolute risk, and a tilt with low tracking error can still sit inside a portfolio that is itself volatile.
Why is the information ratio stated conditionally on the view being correct?
The calculator has no way to judge whether a view will prove right. It computes what the tilt adds if the entered expected returns are realized, and divides that by the tracking error the tilt creates. That produces a ratio of assumed reward to measured risk, which is a way of asking whether the view is worth the deviation it costs. An unconditional information ratio would require a distribution of outcomes rather than a single scenario.
Does a positive return impact mean the tilt is worth taking?
It means the tilt improves the expected return under the assumptions entered, which is close to guaranteed whenever the overweighted sleeve carries the higher assumed return. The informative output is the pairing: how much is added, against how much tracking error is accepted to add it, and against how confident the view actually is. A tilt that adds little while introducing substantial tracking error is the case the calculation is designed to make visible.
Why does tracking error not add up across asset classes?
Because active deviations interact. An overweight in one sleeve and an underweight in another that move together partially offset in the difference series, so the combined tracking error is less than the sum of the individual contributions. Where the two deviations are in assets that move oppositely, the combined figure can exceed what either contributes alone. The per-asset breakdown attributes the total; it does not decompose into independent pieces.
What happens to the result if the view turns out to be wrong?
The return impact reverses sign in roughly proportional magnitude while the tracking error stays the same, because tracking error measures deviation without regard to direction. That asymmetry is the core point of running the calculation before acting: the risk figure is symmetric and known in advance, and the return figure is conditional on a judgment. Entering the opposite expected return for the tilted sleeve shows the downside case explicitly.
How does this relate to a deviation band policy?
A deviation band sets the maximum size a tilt may reach before it must be reduced. This calculation shows what a tilt of a given size does to expected return and tracking error, which is the input to choosing that band in the first place. Running several tilt sizes through the tool converts a band width expressed in percentage points into a tracking error budget, which is the unit most active risk limits are stated in.
Can a tilt be expressed as an underweight to cash?
Yes, and it is one of the more common forms. Reducing the cash sleeve and adding the same amount to another asset class satisfies the zero-sum requirement and expresses a view that being invested is preferable to holding cash. Because the cash sleeve has low volatility, the tracking error such a tilt generates comes almost entirely from the sleeve being increased, which usually makes it larger than a tilt of the same size between two risk assets.
Why do the capital market assumptions matter when the tilt direction is already decided?
They determine both halves of the answer. The expected returns set how much the tilt adds if it works, and the volatilities and correlations set the tracking error it costs. Two people holding the same view can therefore get very different results from the same tilt if their assumptions differ. Changing only the assumptions, with the tilt held constant, shows how much of the conclusion rests on the view and how much on the inputs behind it.