I ended up drawing the Security Market Line by hand one evening because I wanted to see, physically, where a few of my positions landed relative to what the model said they should return. Not because I trust CAPM as gospel. I do not. But the line does something useful that a spreadsheet full of numbers does not. It gives you a single picture where every asset sits somewhere, and the ones that look wrong are visibly wrong. They float above the line or sink below it, and that vertical gap is the whole point.
The idea underneath it is old and simple. The Capital Asset Pricing Model says the return you should expect from an asset depends only on how much it moves with the overall market. That sensitivity is beta. A beta of one means it swings with the market. A beta of two means it swings twice as hard. The model then says higher beta should earn higher expected return, in a straight line, starting from the risk-free rate and passing through the market itself. Plot expected return on the vertical axis and beta on the horizontal, and that straight line is the Security Market Line.
Building the line from real data
You need three inputs and none of them are exotic. The first is the risk-free rate, which is where the line starts on the left. In practice people use a short government bill yield. That is your anchor at beta zero, because something with no market sensitivity should, in theory, just earn the risk-free rate.
The second input is the expected return of the market itself. The market has a beta of one by definition, so it sits at the point where beta equals one. The gap between the market return and the risk-free rate is the equity risk premium, which is the extra you get paid for taking market risk at all. Historically that premium has run somewhere in the mid single digits per year, but honestly the estimate is soft and reasonable people argue about it constantly. I tend to use a modest number rather than an optimistic one, because overstating the premium tilts the whole line and makes everything look cheap.
The third input is beta for each asset you care about. You get this by regressing the asset's returns against the market's returns over some lookback window. Slope of that regression is your beta. A few things to watch here. Use returns, not prices. Pick a window long enough to be stable but short enough to still describe the asset as it is now, so a couple of years of weekly or daily data is a common compromise. And know that beta drifts, so a number you computed a year ago may not describe the same company today.
Once you have those, the line is just a formula. Expected return equals the risk-free rate plus beta times the equity risk premium. Feed in any beta and it hands you the return the model thinks that asset should produce. Draw it across your beta range and you have the line.
Plotting your holdings and reading the gap
Here is where it gets practical. For each holding you now have two numbers. Its beta gives it a horizontal position. And the return you actually expect from it, based on your own forecast, gives it a vertical position. Drop each one as a dot on the same chart as the line.
Now read the vertical distance from each dot to the line.
- A dot sitting above the line is offering more return than its beta says it should. In CAPM language that asset is underpriced. You are being paid more than the model thinks that amount of risk deserves.
- A dot sitting below the line is offering less return than its beta warrants. That asset is overpriced. You are taking market risk and getting shortchanged for it.
- A dot sitting on the line is fairly priced by the model. Nothing to see.
The vertical gap is the signal. In academic terms that gap is alpha, the return an asset delivers beyond what beta explains. A big positive gap is the model waving at you. The key word is expected return, though, and that is entirely your input. The line does not know anything about the asset's future. It only knows beta. You are the one supplying the forecast, so the quality of the whole exercise rests on how honest that forecast is.
Where this quietly breaks
I like the line, but I have watched it lie to people, so a few failure modes are worth naming before you act on any gap.
The obvious one is that your expected return is a guess. If you plug in an optimistic forecast, of course the dot floats above the line. You did not find a mispricing, you found your own optimism plotted back at you. The line is only as good as the number you feed the vertical axis, and it is very easy to fool yourself here.
The second is beta instability. A company that just took on a pile of debt, or pivoted its business, will have a future beta that looks nothing like the one you regressed from the past. Your horizontal position is stale, which slides the dot sideways relative to where the line actually should judge it.
The third is that CAPM assumes a single risk factor, the market, and the world plainly has more than one. Size, value, momentum, quality, and others all move returns in ways a one-factor line cannot see. So an asset can look like it sits above the line purely because it is loaded on a factor CAPM ignores. That is not free money. It is compensation for a risk the model is blind to.
Given all that, I treat the line as a screen, not a verdict. My rough workflow is short. Compute betas on a consistent window so the horizontal axis is apples to apples. Use a conservative equity risk premium so I am not tilting the line in my own favor. Plot everything, then only look hard at the assets with a large vertical gap, ignoring the small ones as noise. And for anything that looks genuinely cheap, ask what CAPM might be missing before I add to it, because a persistent gap usually means the model is blind to something, not that the market is.
When I am doing this across a real book of positions, I lean on the market scorecards and beta tooling inside Blockcircle to keep the betas and return estimates on one consistent footing, mostly so I am not comparing a beta from one window against a forecast built on another. The chart is only trustworthy when both axes are measured the same way for every dot on it.
None of this replaces actually understanding what you own. But as a first pass, sitting down and seeing your whole book spread out against one line, the cheap ones floating up and the expensive ones sagging down, tends to sharpen the questions you ask next. That is usually enough to earn the ten minutes it takes to draw.