Where Your Portfolio's Risk Comes From, in Three Steps

A portfolio of the ten largest US companies, equally weighted, is about the least exotic thing you can own. Ten household names, no leverage, nothing clever.

Run it through our risk model and the biggest single driver of its active risk is Size. Technology sits right behind it, and between them they take close to half of everything that makes that portfolio move differently from the market.

Nobody chose either bet. They came with the construction.

That distance, between the positions you picked and the exposures you’re carrying, is what this tool measures. It’s free, it runs in your browser, and your holdings never leave it.

Open the risk model (free)

What it measures

The model is a weekly cross-sectional factor model of about 3,500 US stocks, micro caps included, estimated over 1,079 weekly cross-sections since 2006 and rebuilt every week.

Every company in it carries a position on 13 style factors (Value, Momentum, Quality, Growth, Stability, Earnings Revisions, Low Beta, Short Interest, Shareholder Yield, Liquidity, Volume Trend, Size and Residual Volatility), plus its industry. The tool matches your holdings to that cross-section by ticker.

So what comes back? Three numbers and a ranking. Total risk is how far the book can move in a year. Universe risk is how far the comparison universe moves. Active risk is how far the book moves differently from that universe, and that is the number describing your decisions rather than the market’s.

How to read it in three steps

Step 1. Paste your holdings

One line per position: a ticker and a weight. Percentages, decimals and dollar amounts all work, and the tool prints which reading it used. Cash goes in as $CASH, and a short position takes a minus sign in front of its weight. You can drop a broker or Portfolio123 export in as it is, with its own header line.

The holdings box of the risk model, showing the ticker and weight format, the run button, and the two example chips underneath it — The whole input. If you would rather not type anything, the two chips at the bottom load an example drawn from the served week.

Step 2. Read the three headline numbers

The three headline readings of the risk model: total risk and active risk in percent a year, and a bar splitting the risk between shared factors and company news — Total risk, active risk, and the split that matters most: how much of the movement is shared with other stocks, and how much belongs to these companies alone.

The bar on the right is the one to read first. It splits the risk into what the shared factors and industries explain and what’s left over for each company on its own. On the ten-name example it lands at 60 percent shared factors and 40 percent company news.

A concentrated book of large names sits about there. But spread the same money across a few hundred positions and the company-specific half shrinks, because single-stock surprises start canceling one another out.

Nothing else in the tool says as much in one glance.

Step 3. Find out what is driving it

The summary panel ranking the top five factors and industries by their share of active risk, with Size first and Technology second, above a table of the five headline risk figures — The ten largest companies, equally weighted, against the All cap universe on the week of 12 September 2026. Size takes 27.4 percent of active risk and Technology 20.0 percent; Growth, Low Beta and Healthcare trail well behind.

This answers the question most portfolios are never asked. Not how much risk, but where it comes from.

So it’s a size and sector bet, with ten tickers sitting on top of it. Whether that’s the bet you want is your call. The job of the tool is to make sure you know you’re making it.

What it will not tell you

It’s a risk description and nothing else. It carries no view on any company, and the only thing it estimates ahead is how far a set of holdings can move. It doesn’t forecast a return.

And it’s candid about its edges, which matters more than it sounds:

The tool names the holdings it can’t cover, drops them, rescales the rest to fill the gap, and tells you by how much. Funds, exchange traded products, non-US listings and companies outside the universe that week aren’t in the model.
When less than 80 percent of the weight you entered is covered, it says so plainly. Below that point, what’s left stops standing for the book you typed.
Where the own risk forecast of a company runs above 150 percent a year, the tool marks the figure, so an extreme number never gets quietly folded into the total.

Open it

The model is rebuilt every week and the page prints the week it came from, so you can tell at a glance how current the numbers are. Simulated equity curves of the classic factors, updated every weekday, are on the factor performance page.

Of everything on that page, the split between shared factors and company news is the reading I come back to most. It tells you whether you’re running a factor bet or a collection of single-stock stories.

The ten-name example takes one click. Whatever you run instead, the useful moment is the same: the screen where the largest driver of your risk turns out to be something you never decided.