Antidote to Election Jitters? CVS Health (CVS)

Images: Courtesy Markus Winkler via Pexels

Hard-to-handicap polls Point to This HealthCare Stock as a Safe Bet

Elections make for the worst of times for polls. As little as six weeks ago, hundreds of approval polls taken since mid-2021 indicated that Kamala Harris was probably the single most unpopular vice president ever. Over that period, more than 50 out of every 100 Americans had a steadily unfavorable opinion of Ms. Harris. And in early July, a full 54% disapproved, while only 37% viewed her in a positive light. How a vice president, with essentially no power, became so actively disliked is an open question.

But disliked she was. 

Kamala Harris was probably the most actively disliked vice president ever tracked. Polling aggregators like fivethirtyeight.com indicated more than 50% of Americans disapproved of her since 2022. But that changed shortly after she emerged as the Democratic presidential candidate.

Then, starting about a month ago, just after Harris emerged as the Democratic presidential nominee, the same models saw her favorability scores jump to about 46 out of every 100. The swing seemed to yield a fungible 3%-ish polling lead over Mr. Trump. The turnaround was apparently confirmed by a fivefold increase in the number of possible winning combinations for Electoral College votes for Ms. Harris, as races tightened for in-contention states like Michigan, Arizona, and Georgia. 

But nothing is ever simple when it comes to election polls. Statistically comparable estimates indicated that both Donald Trump’s favorability and unfavorability scores remained unchanged over the same period, indicating the source of Harris’s new support was not well understood. Wagering platforms that host bets on elections were posting similarly confusing results. Active wagers on one such platform, Polymarket, showed even-money odds for either a Trump or Harris general victory. But when by-state contests were handicapped in the Electoral College, betters wagered that Trump would clearly prevail. 

What gives? These are polls after all. No one is exactly ever sure.

But you can bet on this: Regardless of whom triumphs in November, CVS Health (CVS) will prevail, if only because the company behind this stock – already the world’s second largest healthcare company – controls one the biggest pharmacy benefit outfits in the country, CVS Caremark. Clearly political health and physical health are not correlated.

However, during that same period,  Donald Trump's favorability scores remained unchanged. Explaining what exactly happened requires sophisticated understanding of what polls are, and are not.

Unknowable Polling

Polls have been uncertain things for quite some time. All the way back in 1937, Archibald M. Crossley, an advertising research director, penned the seminal paper on the topic, Straw Polls in 1936. Crossley unpacked how the polls failed to predict Franklin D. Roosevelt landslide victory over Alfred M. Landon the previous year. Crossley laid out the issues for managing an accurate poll. Care is required when capturing the intent of respondents. There are challenges in accounting for geography. There is much work to do to maintain the accurate context for the survey. 

“The greatest difficulty is the fact that the election itself is not a census, but an application of the sampling principle,” Crossley wrote. “Every poll is therefore a sample of a sample.”

The “samples-of-sample” challenges that polls face are not going away. Barbara Bardes, a political scientist, and Robert W. Oldendick, a biochemist, laid out the challenges in their 2012 textbook, Public Opinion: Measuring the American Mind. Here Bardes and Oldendick, worked through what they called the “respondent factors” that cloud polls: Response rates that vary across the means of their collection, say by phone, mail or the Web; how wealth, race, and gender of respondents are factored in. Then comes the devilish details in accounting for the uneven effects for those respondents who knowingly opt in – or out – of a survey. 

In their excellent 2018 paper, Christopher Prosser, a researcher at the University of Manchester and Jonathan Mellon, Associate Professor at United States Military Academy West Point, posit that the effects on a poll by those who opt in or out of a poll is usually not well understood.

“Whether these conditions are met is essentially unknowable, especially before the fact, and the precise combination of relevant conditioning variables will change over time,” wrote Prosser and Mellon. 

So-called respondent factors – that is, how those being polled perceive themselves as they are polled –have always been the challenge in accurate surveys. The evolution of self-identity is making polling more complicated.

The Errors of Omission

Pollsters now face particularly challenging respondent factors. Are respondents identifying themselves as a man or a woman, or are they choosing to be identifiable as a particular gender? Are respondents identifying as members of a racial classification or are they comfortable with their country-of-origin or other geographic marker? Do respondents feel their sense of prosperity is strictly their income less expenses, or is the full value of their property, savings, or even well-being more important?  And when it comes to policy, do respondents identify as a member of a political group based on issues, or are they Republicans or Democrats by convenience or location? 

Pollsters are wrestling with a kind of post-demographic age, where the traditional classifications used to model respondents are becoming less and less effective. What’s intriguing are the upsides that emerge when polls become uncertain. 

The challenges for new the classifications for respondent gender, lineage, wealth, and political affinity make for a fresh perspective on value.

Healthy Bets

It turns out that accepting the uncertainty in election polls makes for a surprisingly effective valuation tool. Let’s examine one of the dog-whistle topics in the November contests: healthcare. There is much speculation on so-called Trump trades in the  business of caring for the sick. The thinking goes that under Republican rule, public sector health resources will be curbed as private sector coverage options will be allowed to flourish. So established providers like UnitedHealth (UNH), Humana (HUM), and CVS Health (CVS) will benefit. UnitedHealth has been considered the most attractive, with company values jumping by about 25% over the past 6 months. Its stock is trading at a 52-week high.

However, if we accept that the upcoming election result is unknowable, we can ask, which of three possible companies will benefit no matter which candidate wins? The winner becomes CVS Health. Why? For the simple reason that it is one of three major administrators for prescription drugs. CVS’s so-called pharmacy benefit management arm, CVS Caremark, along with UnitedHealth’s Optum Rx unit, and Missouri-based Express Scripts manage about 80% of the roughly 3.7 billion prescriptions written each year. 

What grabs headlines about these pharma benefit managers is the bipartisan pressure to regulate them. This month the House Committee on Oversight and Accountability made it clear it will back the Federal Trade Commission attempts to restructure the sector. A round of suits are planned to control prices and protect small-town pharmacies that were being pushed from the market. 

While fines are almost certain, what the actual restructuring to come will look like is a debate. Some say not much will change, as the new regulations meant to protect smaller pharmacies might even increase margins. Long-term, the sense is established pharmacy benefits managers like CVS might see their prospects improve. 

Assuming investors can't know the outcome of the coming elections makes for a surprisingly effective analytics. Beaten-up pharmacy benefit managers like CVS Health will thrive no matter who wins in November.

What makes CVS Health attractive over UnitedHealth therefore is how cheap it is. It's trading right now at a 10.30 price-to-earnings ratio. That is a fraction of the 41 P/E ratio of UnitedHealth. Yet, CVS is among the largest retail pharmacies in the market with 77% of its stock held by institutions. The company yields a not-shabby 4.5% dividend. It’s trading at roughly its book-value per share, at $54. The return-on-equity is a reasonable 11%.

What frightens most investors – and suppresses its value – is CVS’s retail operation. Like many pharmacies it has been closing retail touchpoints by the hundreds. These closures are part of an industry-wide retrenchment in the face of an expanding online medical ecosystem. But those closures won’t affect the fundamentals of the pharma benefits management. 

CVS Health will prosper no matter who wins the elections. Poll or no poll, that’s a prescription for value.