Dear All
I always found the idea of charting price and sales in the log space informative. This is very easy to do in the Fundamental Chart. Just enter the factor, for example SalesTTM, and place it in the same plot as the Price with the Axis set to Independent Log (IL). Here’s what my settings look like
Take a look at two completely different companies (btw, I own both 
)
What the chart does automatically is to fit the series together as best as possible. When you do this with Sales and Price you should see both series hugging each other in the long run. Sometimes the price will be ahead, sometimes behind.
This was the case for FDS, but it got hammered recently because of the idea that AI will destroy SaaS companies. The price now is around 50% less than what the sales trend implies.
For MU it’s entirely different since it is (was?) a cyclical company. The price usually peaked before the sales cycle peaked. The price now seems to be inline with the sales fitted price.
Translating to formulas
I think it’s all very cool for discretionary investors that want to visualize things. But how to translate this to a formula? Basically we want to fit the sales time series into the same units as price. Once we have this “Sales Fitted Price” we can do things like screening for companies trading above or below their fitted price.
There are different methods to do this. ChatGPT offered me two ways: a simple scaling version, and a regression version. I like the regression version better because it gives me a statistically best-fit. For the curious here’s a spreadsheet of both methods.
Here’s a screen that implements the 10Y regression version using our LinRegXY function. As you can see it’s quite complex and took me a while to get it right.
Here are the top 20 companies by MktCap today with their “Fitted Price” and “% from the Fitted Price”:
The most undervalued is LLY at over 52% undervalued vs it’s Fitted Price, with NVDA as a close second. Interestingly MU is overvalued 100% which is different than what the chart showed (the chart is just a simple scaling method)
Where to next?
I like comparing price trends with fundamentals, and I use it visually all the time. I have not used it quantitatively but plan to do so more. For example when I screen for companies well below their fitted price, many show increased insider buying.
The problem with using Fitted Price in formulas is the complexity. It’s time consuming and error prone to write the formula. If there’s interest we could come up with simple factors, for example something pre-built like:
SalesRegressedPrice5Y
SalesRegressedPrice10Y
Or also add more flexible functions with parameters. Would you, for example, want to regress something else other than Sales? Note that since the log space needs to be used, things like EPS will not work due to negative values.
Cheers and Happy Independence Day