Scaling Alpha Forecasts Against Transaction Cost Estimates

Summary

The document presents several proposed ways to set relative objective weights for alpha forecasts and transaction costs in a portfolio optimizer. The setting is monthly standardized factor-model forecasts combined with a nonlinear market-impact model represented by piecewise-linear costs. Suggested approaches include calibrating forecasts against realized portfolio returns, adjusting annual cost estimates for turnover, estimating the relationship at the stock level, or converting standardized scores into return forecasts using information coefficient and volatility.

These are alternatives raised in a question, not a tested recommendation or accepted standard. The examples rely on historical backtest quantities and differ in their treatment of horizon, turnover, and portfolio weighting. No comparative results establish which approach is best, and the document does not specify validation procedures or safeguards against overfitting. Practitioners would need to align units and horizons, then assess calibration out of sample for their own universe, cost model, and portfolio construction process.

Key ideas

• Alpha and transaction cost terms must be expressed on compatible scales in an optimizer.
• One proposal calibrates forecast strength against realized portfolio returns.
• Turnover can be used to adjust annual transaction cost estimates to the holding horizon.
• Stock-level calibration may reduce distortions from portfolio weights.
• The Grinold-style approach converts standardized scores into return forecasts using information coefficient and volatility.
• The document offers candidate methods without evidence that one is universally standard.

Tags

# Is there a standard method of scaling alpha forecasts to t-cost estimates?


# Is there a standard method of scaling alpha forecasts to t-cost estimates?












Given a set of monthly alpha forecasts (i.e. standardized z-scores from a multi-factor return model) and a non-linear market impact model (or more specifically, its piecewise-linear approximation), is there a generally accepted "standard" method of setting their objective weights in an optimization so they are scaled appropriately?

Based on historical returns of a back-test, some of the methods that have been suggested include:

```
Let,

a = average monthly gross portfolio alpha forecast
R = annual realized gross portfolio return
XO = average annual portfolio turnover
```

- Set t-cost objective $= (12 \times a) / R$, set alpha objective $= 1$. This effectively converts the arbitrary z-score to an actual return forecast (or set the alpha objective = the reciprocal of this value, and the t-cost $= 1$).

- Set t-cost objective $= (a / R) * XO$, set alpha objective $= 1$. This treats t-cost as an annual number, and adjusts the holding period horizon based on annual expected turnover.

- Set t-cost objective $=$ the full period average of [stock-level alpha / forward return]. Cross-sectionally more accurate and removes the effect of large portfolio weights on long-term averages.

- Re-scale alphas using Grinold approach: $\alpha = IC \times \text{volatility} \times \text{z-score}$, then objective weight for both alpha and t-cost $= 1$.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.