When a Linear Stock-Price Transformation Adds Analytical Value

Summary

The question asks whether a time series formed as a linear transformation of a stock’s tick-level price series offers any research advantage when it does not lead the original price and has a lagged correlation close to one at small lags. It raises the possibility that any analysis on the transformed series could instead be performed directly on the stock price.

The document does not include an answer or supporting evidence, so it leaves the issue unresolved. In general, a transformation that preserves the same information may not create predictive content by itself, but rescaling or changing representation can still be useful for particular analyses, comparisons, or numerical methods. Whether it adds value depends on the transformation’s properties, the intended task, and whether it makes relevant structure easier to measure. No trading rule or performance test is provided here.

Key ideas

• The question concerns a linear transformation of tick-level stock prices that does not lead the original series.
• High lagged correlation suggests the two series may contain closely related information.
• A transformed series does not establish predictive value merely by being a different representation.
• Rescaling or transformation may still help with specific analytical tasks.
• The document provides no answer, method, or empirical test to resolve the question.

Tags

# Linear Transformation of stock price


# Linear Transformation of stock price












Suppose, using market data for a stock, at a tick level, I arrive at a time series, I(t), which is a linear transform of the stock price time series, S(t). I(t) is not leading S(t) and the lagged correlation between I(t) and S(t) is close to one for a small t.

My question is that for such a case, does the researcher achieve anything at all? Any analysis which can be performed on I(t) can be performed on S(t) and since I(t) doesnt lead, we dont have any advantage at all.

Not sure if what I have said makes sense but would like to know experts' opinion on the same.

Regards

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.