Incremental Returns and Expectations

Stock Picking in a Nutshell

Picking stocks is about expectations. Most of the time, a stock is “fairly” priced given the information available. Stocks move when surprises happen and investors update their expectations for a company.

Of course there are lots of other factors driving stock prices: economy-wide macro events, interest rates, short term sector moves, government policy, and so on.

But essentially, it’s best to own a stock whose surprises are more likely to be positive than negative.

When expectations are low, even small positive surprises can lead to a good return. If the market expects a terrible result, even a mediocre outcome will move the stock up.

On the flip side, stocks are most dangerous when expectations are very high. It’s hard to have a positive surprise when consensus thinks “the sky’s the limit” already.

Getting a superior investment return depends on having a variant view from consensus, and then being right. Where do variant views come from? It helps to have a longer-term perspective on the market, understand how investors form their beliefs about companies, and take a bigger picture view of popular investment themes of the time.

Historical Perspective: Evolutions in Investment Strategy

Robert Hagstrom (best known for his detailed unpacking of Warren Buffett’s investment methods) describes the evolution of five different investment strategies and the eras when they were dominant. Paraphrasing from his book, these are the major stock selection philosophies of the past century.¹

1. 1930s-1940s: discount-to-book value strategy (“net-nets”) proposed by Ben Graham and David Dodd in Security Analysis.

   Buy companies whose assets are worth more than the market value of the company’s shares. A conservative, defensive approach, suited to the post-Great Depression era where a business “worth more dead than alive” was an appealing backdrop.

2. Post-WW2: Dividends. Buy stocks that pay a high dividend yield.

   Over time, the dividend yield on stocks became lower than the yield for bonds (historically unusual!) as rising prosperity made long-term investing in stocks seem increasingly safe.

3. Go-Go 1960s: choose stocks whose earnings are expected to grow at a high rate.

   The focus on “growth at any price” led to the Nifty Fifty (e.g. McDonald’s, Xerox, Polaroid, Avon) whose overvalued shares crashed in the bear market of the 1970s.

4. 1980s: select for high “owner earnings” or cash flows relative to the stock price.

   Some of Warren Buffett’s iconic investments happened here; stocks like Coca-Cola, whose shares had become cheap but still had excellent fundamentals.

5. Most recently: Cash Return on Invested Capital a.k.a. “quality” investing.

   Best exemplified by Terry Smith, whose Fundsmith Equity Fund had a fantastic run up until roughly 2022. He emphasizes asset-light businesses, with strong intangible advantages (brands, network effects, intellectual property) that require little cash reinvestment to grow.

Finally, deserving a spot on this list as well, is passive investment through index funds. Don’t search for the needle, own the entire haystack has become an increasingly popular model for no-hassle investing.

Every investment philosophy that works carries some element of truth. We should indeed care about the assets on a company’s balance sheet if they hit hard times; stocks that pay steady dividends do have something to recommend them; and over long holding periods, earnings growth and stock performance do track together.²

The trouble for any approach is that when it’s working, it spawns imitators. Expectations get higher and disappointments become more likely. Each era of investing carries the seeds of its own obsolescence, and so we can’t afford to stay stuck in models just because they’ve worked in the past.

When the Best Companies Aren’t the Best Investments

Early in my investing career I found the “quality” approach highly compelling. Study financial statements carefully and own only the very best, most profitable companies. A business that can internally invest very efficiently, generating high returns on capital, should logically be a superior long term investment for the stockholder as well.

But what happens when everyone else in the market thinks the same thing?

So-called “quality compounder” stocks became a crowded trade. Then as generative AI tools rapidly improved, the competitive advantage behind many businesses started to look shallower and investment returns in that style have since suffered.

Crucially, it’s not sufficient to identify a great business. That sort of research has gotten easier to do than ever before. Rather, we want to spot good businesses that have the potential to become great (or at the very least, marginally better).

Recent research by Michael Mauboussin and Dan Callahan (2023) supports this approach. Their paper splits companies into quintiles, where the best are the top 20% by returns on invested capital (ROIC), the worst are the bottom 20%, and the rest are divided in between. The paper has more points to make but here I’m reproducing only their most provocative results.³

To read this table: suppose you invested only in top-quintile (very high ROIC) companies. At the end of 3 years, if they were still among the top, the average total shareholder return (TSR) was 20% per year. However, if a top company fell to the 3rd quintile (still above average ROIC, but not exceptional) then over 3 years the investor earned 0%!

Key point: The direction of change matters more than the absolute level of profitability that a firm exhibits.

But how likely is each of those outcomes?

Within this sample, 48% of top quintile firms were still exceptionally high ROIC after 3 years and 41% of the worst firms were still the worst. Clearly there is persistence of quality over time, but it’s far from a sure thing.

What we care about as investors is our chance of each outcome, multiplied by the return we’d receive in each case. Here is that table, combining the information in the previous two:

The contrarian finding is that stocks in the middle of the pack (the 2nd, 3rd, and 4th quintiles) outperform the best (“quality”) stocks as well as the worst (“turnarounds”). While the best companies that stay the best give good returns, if they slip even slightly then investors are punished. Middle-quality stocks have less divergence of outcomes, and are also strongly rewarded for improvements on the margin.

This is only one study, but I take it seriously because it matches intuition: expectations matter, and if everyone can spot the “best” companies on ROIC then that ceases to be a good way to predict outsize stock returns. If anything, it becomes a slight headwind.

A few qualifiers:

1. The study assumes a 3-year minimum holding period. This is not a short term trading strategy.
2. No stock-picking skill is assumed, just holding all stocks of each type for 3 years. Could results conceivably be improved through careful study and selection? Maybe.
3. The study used data from 1990-2022, before commercial generative AI hit the market. It seems likely that rates of change within businesses are getting faster and persistence within the same quintile is less likely going forward, although we can’t know that for sure.

This result shapes my current investment process. Can we identify companies that are stable but unexciting now, showing adequate but not exceptional ROIC, where market expectations might be low? And can we find reasons such as favorable economic trends, intelligent management decisions or some other sort of self-help, to indicate that the business could surprise to the upside over the next few years?

Researching to spot those rare opportunities is where I dedicate my time.

Important Information

This article is general educational information and reflects the author’s views as of November 2026. It is not a recommendation to buy or sell any security, a solicitation, or personalized investment, tax or legal advice. No advisory relationship is formed until a written agreement is signed by both parties.

Third-party research: Exhibits 9 and 10 reproduce data from Michael Mauboussin and Dan Callahan, “ROIC and the Investment Process” (Morgan Stanley Investment Management, 2023), based on FactSet and Counterpoint Global data for 1990–2022, excluding financials and real estate. Bormann Wealth Management is not affiliated with Morgan Stanley or Counterpoint Global and has not independently verified the underlying data. Investors, funds and companies named in this article, including Fundsmith, Coca-Cola and the “Nifty Fifty” stocks, are historical illustrations of investment approaches, not recommendations.

Hypothetical calculations: The expected TSR table is a hypothetical illustration prepared by Bormann Wealth Management by multiplying the published transition rates by the published average returns. It does not represent the performance of any client account or strategy, assumes no stock selection, and does not reflect fees, taxes or trading costs. A probability-weighted average of annualized returns is not a compounded portfolio return, and averages of this kind may cover a wide range of outcomes for individual stocks, including losses.

Conflict of interest and forward-looking views: This article describes the research approach Bormann Wealth Management uses to select stocks for clients, so the firm has an incentive to present it favorably. Selecting individual stocks based on expected changes in profitability may underperform the broad market, and a portfolio of individual stocks can be more volatile than a diversified fund or index. Statements about artificial intelligence and future rates of business change are the author’s opinions and may prove wrong.

Investment risk: All investing involves risk, including the possible loss of principal. Past performance is no guarantee of future results.

Positions and scope: As of the last review date, the author, Nick Bormann, does not hold shares of companies named in this article in his personal accounts, and may buy or sell any security discussed at any time without notice. Bormann Wealth Management, LLC is an investment adviser registered in the State of Washington; registration does not imply a certain level of skill or training. See Disclosures and Form ADV Part 2A.