Perfect Signal Analysis

AKAnomics is focused on building a robust process for estimating revenues of companies. Its current focus is on companies that are highly sensitive to the gyrations in the global economy (see Power of Granular Nowcasting). A question that clients often ask is “Why Revenues” & “Why not some other metrics like margins”? Our quick answer is – Revenues are a highly relevant indicator for fundamental analysis, and good estimates of revenues can yield significant value for investors.

The two major candidates for fundamental indicators for company performance would be either revenues or margins, and we have geared our process towards revenues. Changes in margins are far more sensitive to changes in revenues than changes in cost, especially for revenues of economically sensitive businesses. Even during periods of rapid commodity inflation, company costs are often dampened by a large component of cost that is fixed. Even among companies with a large component of variable input/commodity costs, these cost changes could still be slower relative to the revenue changes from the economy. And finally for many companies (e.g., Chemical/Transportation companies) that are highly sensitive to commodity costs, the companies tend to pass on some costs to their customers, making revenues even more relevant.

The fundamental arguments towards picking revenues as a key indicator, can be further supplemented with the potential value that can be derived from having advanced perfect knowledge of revenues. While it is impossible to have such information, it nevertheless provides us with an upper bound of the value, which can be further used as a reference for a “less than perfect” practical methodology. For this value analysis, we simulate a long/short strategy that trades each week the names where the (perfect) revenues for the upcoming quarter for the 118 companies we track are above/below consensus beyond +/- 0.75 times consensus historical standard error. The chart below shows the historical cumulative paper sector-neutral returns (relative to the Industrials Index called XLI) for such a strategy with an impressive Sharpe Ratio of 3.3, confirming our thesis of revenues being a valuable indicator to target for our methodology.

A similar experiment on historical AKAnomics models suggests a Sharpe Ratio between 1 and 1.5 depending upon which sub-sectors are included/excluded. In addition to the value our process could create for systematic and discretionary investors, there is also room for improving the process further.