Factor Investing: What It Is and When It Stops Working

Factor investing means buying stocks for a measurable trait instead of a forecast. The trait is called a factor. Cheapness and recent price strength are two of them. Each has historically explained why one basket of stocks beat another.

The rule does the picking. Nobody sits in a room deciding they like the CEO.

What is factor investing, in plain terms?

Start with a universe of stocks. Score every name on one measurable trait. Buy the names that score highest. Rebalance on a fixed schedule and repeat.

That's it. The whole discipline is one rule, applied without discretion, for a long time.

The academic spine came from Eugene Fama and Kenneth French. They published the three-factor model in 1992, adding size and value to plain market exposure. They extended it to five factors in 2015 with profitability and investment. Practitioners added momentum and low volatility along the way.

How does factor investing actually work, step by step?

1. Define the universe. Usually a broad index like the S&P 500 or a global developed-market set.
2. Define the metric. Book-to-price for value. Trailing twelve-month price change for momentum. Return on equity and debt levels for quality.
3. Score and rank. Every name in the universe gets a number.
4. Select and weight. Take the top decile or quintile. Weight by score, by market cap, or equally.
5. Rebalance on a calendar. Monthly for momentum, annually for value. Faster rebalancing costs more to trade.

Notice what's missing. There's no step where the rule asks what the market is doing. A value screen in March 2020 looked exactly like a value screen in March 2019.

What are the main factors an advisor will run into?

• Value. Stocks that are cheap against book value, earnings, or cash flow.
• Size. Smaller companies over larger ones.
• Momentum. Names that have gone up recently.
• Quality. Profitable companies with low debt and stable earnings.
• Low volatility. Stocks whose prices move less than the market's.

Each one has decades of published research behind it. Each one has also spent long stretches losing money relative to a plain index fund. Both things are true at once.

What does a factor failure actually look like?

Momentum in 2009 is the cleanest example on record, and the dates matter.

The S&P 500 closed at 676.53 on March 9, 2009. That was the bottom. A momentum rule on that date was doing precisely what it was built to do. It ranked names by trailing price strength, so it held defensive stocks and staples. It had almost nothing in banks, homebuilders, or anything else that had been destroyed over the prior eighteen months.

Then the market turned, and the most damaged names rose hardest. By the end of June 2009 the S&P 500 had climbed back near 919. The names momentum had ranked at the very bottom led that move. Momentum posted one of its worst stretches in decades, in a quarter when the index went straight up.

The signal wasn't broken. It fired correctly on a trailing window. A sharp reversal is exactly what a trailing window can't see.

Value ran the same trap for far longer. From 2007 through 2020, value trailed growth across most of thirteen years. An advisor who bought a value fund in 2007 spent over a decade explaining it. Then 2022 arrived. The S&P 500 peaked at 4,796.56 on January 3, 2022. It closed at 3,577.03 on October 12, down about 25 percent. Value held up far better than growth through that year.

The factor worked. The client who sold in 2020 wasn't there for it.

When does factor investing stop working?

Four failure modes come up repeatedly, and only one of them is about the research being wrong.

The holding period is longer than the client. Factor premiums are measured over decades. Advisory relationships get judged over quarters. A thirteen-year drought is survivable on a spreadsheet and not always survivable in a review meeting.

Crowding shrinks the premium. When a factor gets popular, more money chases the same names, and prices adjust before the rule ever fires. Some of the published premium was the reward for owning something uncomfortable. Comfort is expensive.

Every provider defines the factor differently. Two quality funds can share a label and hold almost nothing in common. One screens on return on equity, another on earnings stability, another on accruals. Read the methodology document, not the fact sheet.

The rule answers the wrong question in a drawdown. A factor screen tells you what to own inside the equity sleeve. It never tells you how much equity to own. In 2008 and again in early 2020, every long-only equity factor fell hard. They were all fully invested by construction.

That last one is the part advisors underestimate. Factor selection and risk posture are two separate decisions, and a factor fund only makes the first one.

How is factor investing different from smart beta?

Mostly they're the same idea under two names. Factor investing is the academic term for the underlying effect. Smart beta is the marketing term that got attached to the products. It drifted until it covered almost anything that wasn't cap-weighted. We wrote about how that happened in our piece on smart beta.

When you see either label, go to the methodology. The label tells you nothing about what the fund holds.

Where a systematic risk overlay fits

THOR's work sits next to factor investing, not inside it. A factor rule scores what to own. An adaptive model scores whether to be invested at all, and shifts exposure when trend and volatility conditions deteriorate. The approach seeks to reduce drawdowns rather than pick the cheaper stocks.

Advisors often run both. A factor tilt inside the equity sleeve, and a systematic overlay deciding the size of that sleeve. The low volatility factor is where the two overlap directly. Lower-volatility names tend to fall less in a selloff on their own.

Definitions

• Factor. A measurable stock characteristic that has historically explained differences in return across a universe.
• Factor premium. The excess return a factor has delivered over a broad market index across a long measurement period.
• Fama-French three-factor model. A 1992 framework explaining returns through market exposure, company size, and value.
• Momentum. A factor that ranks stocks by trailing price strength. The window is usually twelve months, skipping the most recent one.
• Crowding. The condition where wide adoption of a factor bids up its target names and compresses the premium.
• Single-factor fund. A fund built on one screen. A multi-factor fund blends several, which smooths results and dilutes each one.

Fama and French never argued a factor would work every year. The 1992 paper measures a premium across decades, and the drawdowns inside those decades are in the same data.