S&P, Moody’s, Fitch, and the other credit rating agencies analyze fixed income securities to categorize them by probability of default. Their ratings reduce the risk of rated securities by increasing investor knowledge and thereby acting as a standard in the capital market supply chain.

Yet the rating agencies have an inherent conflict of interest. They’re paid by the firms that they’re supposed to be objectively analyzing. The creditor firms can shop around for the highest rating, incentivizing the rating agencies to improperly inflate ratings to attract customers and win market share.

Everyone knows this. Steve Carell learned it in The Big Short when he tried to confront S&P about their ratings and the representative told him that if they don’t give out high ratings then their customers will take their business to Moody’s.

The Wall Street Journal frequently reminds its readers. They published an investigation in August that showed “S&P and its competitors are giving increasingly optimistic ratings as they fight for market share.”

WSJ journalist Cezary Podkul was on CNBC to explain:

Mathisen: Is the problem as acute as it was pre-financial crisis? A lot of people, as we said in the introduction, thought this was a problem that had been hosed out of the system. But clearly it hasn’t.

Podkul: No because it’s the same business model. The main thing that congress didn’t change after the financial crisis is they said, ‘hey issuers, you can still pay for the ratings.’ And so that gives them leverage to decide who to pick. And, of course, they have an incentive to pick the best ratings. So because that wasn’t fixed, that whole dynamic is still there.”

Elizabeth Warren smelled blood in the water. She sent a letter to the SEC in September demanding to know why they haven’t forced the agencies to change from the ‘issuer-pays model’ to the ‘subscriber-pays model’.

In the subscriber-pays model, investors pay the rating agencies for the ratings. Supposedly, a subscriber-pays model will solve the conflict of interest because creditors no longer have the opportunity to shop around for ratings.

Yet even in this new structure the industry would run into the same problem. Matt Levine at Bloomberg makes the point:

“Surely sometimes the ratings firm would decide that a bond was a double-B, and some big asset manager would call it and say “we really like that bond and we want to put it in our investment-grade fund, can you make it a triple-B?” They might not say it like that; they might make a credit argument instead (presumably they like the bond because they think it’s a good credit!), or they might just go shop for a better rating from a different ratings firm.”

Alignment of Interest

Maybe the problem is not with the agencies’ business model but with how people think about the business model.

Does the credit supply chain above look different than any other supply chain?

The same argument about a conflict of interest applies to car companies. The automotive OEM can shop around for the seatbelts that go into its car. The OEM wants the lowest priced seatbelt and thus seatbelt suppliers are incentivized to offer the cheapest, lowest quality seatbelts to gain market share. This dangerous conflict of interest will result in seatbelts made of tissue paper!

Why doesn’t this happen? Because the OEM doesn’t want the cheapest, lowest quality seatbelts- it wants the seatbelts that have the highest net value to the customer.

The only reason that seatbelts have value to the OEM is because they have value to the dealer. And the only reason they have value to the dealer is because they have value to the car owner. The same is true of credit ratings.

The creditor doesn’t want the highest credit rating on their debt as Podkul claims- it wants the credit rating that has the highest net value to the investor and thus most lowers the cost of its debt. The only reason that credit ratings have value to creditors is because they have value to asset managers. And the only reason that the ratings have value to asset managers is because they have value to individual investors who benefit from the proper classification of their investments’ risks.

What you might expect to happen in a properly functioning market is that inflated ratings are less valuable to the investor, and thus less valuable to the creditor.

That is exactly what the WSJ reports: “Some bonds in markets where ratings criteria have been eased don’t trade at the high bond prices their ratings suggest they should. Investors have also shown skepticism about ratings on some corporate and government bonds.”

This is one reason why it’s hard for competitors to take market share in this industry. You can’t simply lower the quality of your product and vacuum up market share.

Conflicts of Interest

Of course, a well trusted rating agency could inflate ratings to gain market share before investors realize what they are doing. A rating agency could spend its reputational capital for short term gain just like an insurance company can grow tremendously by selling unprofitable insurance. It turns out that short-term profits at the expense of your long-term business is an impractical strategy for making money. Every business that tries it arrives at the long-term part of the equation. We should be glad that Elizabeth Warren isn’t in charge of strategy at one of the agencies.

In essence, the arguments against the rating agency business model amount to: ‘it’s possible for an agency to do something that’s profitable in the short run but disastrous in the long run’. But that’s true of every choice in life. That’s why it’s possible to invent a ‘conflict of interest’ in any business model that’s dreamed up.

Everyone has a thousand opportunities a day to make choices with short term payoffs and awful long-term consequences, but we properly conceptualize those as ‘self-destructive’. You don’t face a conflict of interest between saving for retirement and blowing your savings on drugs for momentary enjoyment.

Yet people often make decisions that sacrifice the long run for immediate gratification. Some people do steal, or lie to their significant other, or whatever. Suppose you didn’t want someone to engage in self-destructive behavior. Would you frequently remind them that the behavior is in their interest?

“Of course, the truth of our generalizations is dependent upon the validity of our concepts. An invalid concept is a red light to induction; it stops the discovery process or actively leads to false generalizations.

Recall the concepts of “natural” and “violent” motion in Greek physics. These concepts made a fundamental distinction between motions that are in fact similar; for example, smoke rising in air is regarded as “natural” but wood rising in water is regarded as “violent,” and a ball swung around on the end of a rope is moving “violently” but the moon orbiting Earth is moving “naturally.” On the other hand, they group together motions that are very different; for example, rising smoke and falling rocks are both moving “naturally,” and a ball swung in circles and another ball with constant horizontal velocity are both said to move “violently.”

Such concepts cannot be reduced back to observed similarities and differences. They are juxtapositions rather than valid integrations, and therefore it is impossible to reach true generalizations among their dissimilar referents. When the Greeks tried to generalize, they were led into a series of falsehoods.”

The Logical Leap: Induction in Physics – David Harriman

Like the ancient Greeks, the modern investment industry has a debilitating conceptualization problem stemming from historical accident and perpetuated by confusing terminology.

Last weekend’s WSJ article by Jason Zweig is the latest to bemoan the decade-plus relative underperformance of value stocks relative to growth stocks.

Zweig cites a recent study by Rob Arnott at Research Affiliates. The study’s conclusion focuses on ‘bad luck’ as the cause of the relative underperformance and suggests that the expected returns of value strategies are as close to as attractive as they’ve ever been.

Yet as Warren Buffett has pointed out countless times, growth is part of valuation. From Buffett’s 1992 letter to shareholders: “Growth is always a component in the calculation of value, constituting a variable whose importance can range from negligible to enormous and whose impact can be negative as well as positive.”

Growth is not strictly an input into valuation except by the convenience of the valuer. If one knew the future cash flows of a security (as for a fixed income instrument) one would simply input those precise cashflows into the discounted cash flow model. The growth of those cash flows, a relative comparison from one to cashflow to the next, is a descriptive element. It’s only because of convenience that growth is usually chosen as a driver in DCF models. Growth is a component of value like height is a component of a person.

Note Zweig’s implicit perspective on the growth-value issue: “value stocks should eventually outperform simply because their shares are cheaper.”

Just as the Greek’s use of ‘violent’ motion to conceptualize dissimilar phenomenon led them into a series of falsehoods, so the term ‘cheap’ leads many investors into the same trap. The concept ‘cheap’ is an invalid package deal- a misintegration of fundamentally dissimilar items that are grouped together because of an unimportant similarity. Both securities that are priced low relative to their immediate cash flow and securities that are priced low relative to their intrinsic value are subsumed by the concept despite being very different phenomenon.

There is a superficial similarity between the two types of ‘cheap’ in that they both designate that a security is priced low relative to some standard, but those standards are not the same and shouldn’t be equated. Zweig uses one kind of cheap in one context, and the other kind of cheap in a separate context, and then draws a conclusion based on their false equivalency.

The word ‘cheap’ is used both ways in common parlance, but context generally makes clear what the speaker means. When John says that his used jalopy was really cheap, it’s understood that he probably means that its cost was a low absolute dollar amount. When Sally says she got her new Lamborghini for really cheap, she probably means that it was inexpensive relative to the value of the car and not a low absolute dollar amount. This works fine for everyday life, but the investment industry requires more precise concepts than the man on the street. Imagine the state of physical science if it had not moved past the concept of ‘weight’ to the more precise concepts of ‘mass’ and ‘acceleration’.

Unfortunately, the confusion doesn’t stop at ‘cheap’.

Ben Graham’s contributions to the science of investing were indispensable. Intrinsic value, margin of safety, a focus on fundamental analysis, and the Mr. Market metaphor were tremendous contributions to the field. His practice of earning excess returns by purchasing securities below their intrinsic value came to be known as value investing. Graham also identified a repeatable process for identifying such securities. By selecting among low-multiple stocks, investors had a superior chance of finding securities trading below their intrinsic value. The process was so powerful that a generation of investors was able to find great success by implementing a similar process. By selecting low price to book stocks out of Value Line, Walter Schloss beat the market by an astounding 12.9% a year for over 28 years. That was despite owning over 100 positions at any given time.

Certainly, there were some high-multiple stocks that were undervalued in Ben Graham’s time, and in some other universe it could have been that higher multiple stocks was the more lucrative pond to fish in. But for whatever reason (psychology? financial ignorance? institutional imperative? Non-disruptive period in the economy?), it was the low multiple stocks that contained more than their fair share of undervalued securities. Or at least, it was the low multiple stocks that were easier to identify for quantitative reasons. Value Investing developed with a deep connection to low multiple stocks, which became known as “value stocks”. By sharing the word ‘value’, ‘value investing’ and ‘value stocks’ are victims of the same conceptual problem as ‘cheap’. The undervaluation of a security and a low multiple are two different phenomenon that need to be separated to think clearly.

 ‘Value Stocks’

Fama and French argue that the outperformance of the Value Factor (ie. low multiple stocks) is compensation for the higher risk of those stocks. That seems theoretically plausible, yet many in the traditional value camp claim that the lower multiple stocks are actually safer because you are paying less for a piece of the future. Applying the claim to fixed income reveals how little sense this makes. Would anyone claim that high yield bonds are safer than treasuries because the yield is high?

And like treasuries, some common stocks trade at low yields/high multiples for the exact reason that they are less risky. Utilities and consumer staples have often traded at higher multiples despite low growth because their cashflows are predictable. You can characterize low multiple stocks as ‘paying less for a piece of the future’, as Zweig does, but that future may be far worse than you predicted.

While the future for many growth stocks is difficult to predict, there are also growth companies with greater cash flow predictability than some no-growth companies. In other words, there’s nothing intrinsically unpredictable about growth. Macy’s revenue hasn’t changed much over the last five years while Visa’s revenue has almost doubled. Yet it’s far easier to have confidence around Visa’s 3 year forward cash flows than Macy’s 3 year forward cash flows. Companies dependent on a commodity price are also frequently slow growing businesses that are difficult to predict.

The Future of the Value Factor

There are only two reasons that low multiple stocks could have outperformed historically. One is that they outperformed because they deserved to, or in other words, because they were properly valued but riskier. The other is that they outperformed despite not deserving to, or in other words because they were undervalued.

If the former is the case, it seems likely that we will see a return to the long-term outperformance of low multiple stocks as compensation for additional risk.

If the latter is the case, it seems likely that the systematic undervaluation is gone (or leaving) as a durable phenomenon. It’s very hard to imagine that 150 years from now there will be investors buying low multiple stocks and outperforming everyone else with the same or lesser risk.

It feels like inauspicious timing to be making a conceptual case for higher multiple stocks since valuations do appear stretched, particularly in specific corners of the market. It also feels like we’re in a period where an appreciation of risk isn’t being properly compensated for. And none of this is to say that the low multiple stocks won’t outperform higher multiple stocks in the immediate future or even over the long-term, but it is to say that nothing of importance turns on that happening. Buying securities for less than they’re worth will always work eventually.