Simulating Valuation Ratios for Investment Strategy Tests

Summary

The document raises a problem in simulation-based testing of investment strategies that use valuation measures such as price-to-book or dividend yield. A simulation that generates prices while retaining historical values for accounting variables may create unrealistic combinations, because market prices and financial statement measures are related. The author also questions whether correlations between these variables can be estimated reliably from quarterly observations, making a simple scaling approach seem unsuitable.

No simulation design, strategy result, or literature reference is supplied; the text is a request for methods. Its main research consideration is therefore the joint modeling of market and accounting variables while accounting for their dependence and the limited frequency of financial statement data. Any proposed approach would need to preserve economically plausible relationships and address uncertainty in those estimates. The document does not specify an asset universe, strategy, or evaluation criteria, so it does not support a particular modeling recommendation.

Key ideas

• Strategy simulations using valuation ratios need to account for dependence between prices and financial statement variables.
• Holding accounting variables at historical values while simulating prices can produce implausible ratios.
• Quarterly observations may provide an unstable basis for estimating relationships between market and accounting data.
• The document identifies the modeling challenge but offers no method or empirical results.

Tags

# Simulation of price ratios


# Simulation of price ratios












How to go about simulations of variables like price-to-book or dividend yield? Basically I would like to do a simulation based testing of an investing strategy (other than historical simulation). It’s doesn’t make sense just simulating price and taking the other variables from history because they’re obviously very dependent, also the correlation estimate of price and financial statement variables would be pretty unstable since it’s quarterly data, so something like Beta scaling of the variables seems like a bad idea.

Any ideas or literature references for this?

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.