Bristol-Myers Squibb: The Opdivo Fiasco

| About: Bristol-Myers Squibb (BMY)


BMS surprised the Street again with disappointing subgroup analysis of its important study of Opdivo in first-line non-small cell lung cancer.

The stock is now down more than 33% since its recent high.

Competitive forces for Opdivo in the PD-1 market are discussed.

Then, the implications of such a large stock market crash for a major company such as BMS are discussed at length.

A cautious view of US equities is re-examined and reaffirmed.

Bristol-Myers Squibb gets bloodied in lung cancer

First, the news and implications for investors of the disappointing details for Bristol-Myers Squibb (NYSE:BMY) investors with the presentation, then the larger implications.

In early August, BMY released top-line data from one of its many studies involving its checkpoint inhibitor Opdivo, an immuno-oncology (I/O) antibody indicated for several cancers.

This PD-1 inhibitor is in competition with Keytruda from Merck (NYSE:MRK). Both Opdivo and Keytruda are approved for second-line treatment of the predominant form of lung cancer, NSCLC. However, Keytruda has followed a personalized medicine path, and MRK had announced success in a Phase 3 trial of Keytruda in a minority subset of NSCLC patients - those with at least a 50% rate of PD-L1 expression by cancer cells. That subset involved patients with a high level of the protein (PD-L1) to which PD-1 binds.

BMY has followed a different path with Opdivo that as of now has it in the sales lead over Keytruda. The BMY approach was to study Opdivo in NSCLC patients without regard to their PD-L1 level. That worked well until the release in early August that in contrast to the much more targeted approach by MRK, Opdivo failed to show benefit in first-line disease when compared to standard chemo.

There was substantial hope that when the details were released this week at the European Society for Medical Oncology, the Opdivo data would show that the high PD-L1 group would show benefit, with the subgroup with low levels being the group with no benefit.

No such luck.

As BMS reported Sunday:

The topline results from this study were previously disclosed and showed CheckMate -026 did not meet the primary endpoint of superior PFS compared to chemotherapy.

To make things more readable, I'll now quote from FiercePharma about the subgroups:

Among patients with PD-L1 expression levels of 5% or greater, Opdivo failed to top chemo in progression-free survival--posting a median 4.2 months versus platinum-based doublet chemotherapy's 5.9...

Note this is only a 5% PD-L1 level. The article continues with what really hurt:

Breaking it down by subgroup [e.g. 50% PD-L1 cut-off level] also did not yield any Opdivo benefits. "We have not seen major differences in the biomarker cutoffs," Fouad Namouni, Bristol's head of oncology development, told FiercePharma in an interview at ESMO.

However, as so often occurs, there's somewhat of a "but," and note where my emphasis is added:

Namouni did note some imbalances between the study's two arms, including more patients with high-level PD-L1 expression in the chemo arm. "It makes the interpretation of the data at the high level of cutoffs very difficult because it was not stratified," he said.

So does that mean that BMY can redo a study using a 50% cut-off level and catch up? Maybe not, as the article continues (not quoted), saying the BMY is not going to redo a solo study but instead focus on combinations, such as with Yervoy or chemo.

I'm no oncologist, but from my perch as a retired clinical cardiologist, I'm not loving this strategy. So, even though I was looking to buy BMY at $50, or roughly 20X GAAP estimated 2016 EPS, I passed. (Previously I had flipped BMY when it fell to around $55, so I had only held it very briefly.)

BMY may have been too clever by half here, which casts doubt on whether a 20X P/E is especially attractive given all the I/O competition coming (see below) and the boost for Keytruda in at least first-line NSCLC.

MRK and others may lead forward in I/O

This looks like a big win for MRK, which simultaneously released detailed data on Keytruda in this first-line setting at ESMO along with a prestigious article in the NEJM.

MRK does not have a lot of growth drivers beyond Keytruda, and with its annual revenue near $40 B, it's difficult even for a presumptive mega-blockbuster such as Keytruda to impress me given MRK's $174 B market cap at a price around $64 per share.

That's partly because of the following positive news from a PD-1 competitor that has not gotten nearly as much publicity as Keytruda and Opdivo. That competitor is Tecentriq, a Roche (OTCQX:RHHBY) antibody; as described by FiercePharma:

'Unprecedented' Tecentriq data sets stage for Roche lung cancer battle with Merck, BMS

Unprecedented could be very good beyond this one study. Here are some details from that article:

Get ready to move over, Merck and Bristol-Myers Squibb. Roche has some new Tecentriq data that could soon help it break into the lung cancer space--and give its immuno-oncology rivals a run for their money.

In a Phase III study presented at the ESMO 2016 Congress, the Swiss pharma giant's PD-L1 checkpoint inhibitor improved overall survival among previously treated non-small cell lung cancer patients by a median 4.2 months over chemo. On Tecentriq, patients lived a median 13.8 months, topping the 9.6-month figure for the chemo arm.

The study enrolled patients who'd failed on one or two previous treatment regimens, and regardless of whether they expressed the PD-L1 biomarker.

Importantly, Roche is pretty darn bullish here:

That 13.8 month median survival figure is "essentially unprecedented," Dan Chen, head of cancer immunotherapy development for Roche's Genentech unit, told FiercePharma in an interview at ESMO. "It suggests that Tecentriq really does have a powerful survival effect."

As stated above, this could be a strong marketing tool against Keytruda and Opdivo in second-line NSCLC.

The study may shed light on Opdivo's failure, as the article continues:

That survival benefit was "pretty similar" among those patients who expressed PD-L1 and those who didn't, Chen said, suggesting that "not only do you get a benefit in patients who have preexisting immunity," but patients who don't "have evidence of preexisting immunity can get a survival benefit" as well.

That's an "on the one hand" part of the article. It supports the BMY approach in second-line disease. However, the article continues and points the following out:

On the other end of the spectrum, however, patients with the highest PD-L1 levels seemed to derive a greater benefit, posting median survival of 20.5 months in the Tecentriq arm compared with just 8.9 months in the chemo arm.

So, degree of PD-L1 does matter, even in second-line disease.

The BLA for this indication has already been submitted; Tecentriq is already approved for bladder cancer.

One reason I'm cautious on BMY as well as MRK is this Tecentriq data, in conjunction with a giant program that Genentech/Roche has for this drug.

But there's more competition coming. Most notable is durvalumab from AstraZeneca (NYSE:AZN), which is in Phase 3 for a number of indications. AZN has partnered with Celgene (NASDAQ:CELG) for combo use.

Among companies that are farther behind are Regeneron (NASDAQ:REGN) supported by Sanofi (NYSE:SNY) in the I/O space. A potentially registrational study in skin cancer from their candidate is in the early stages.

If Amgen (NASDAQ:AMGN) is competing in this space, its candidate is not in Phase 3. (I think it should have been a leader here rather than chasing CV drugs. Where's its focus on core competencies and giant, expanding markets?)

In any case, there's plenty of competition, and likely more coming; as well as from alternative therapies.

I'd like to do an interim summary, then move on to a broader topic.

Interim summary - I/O a tough space

Neither BMY nor MRK has really great scale in I/O. Neither does AZN; and, as of now, neither does CELG - it has the IMiDs and that's that.

Whereas, Genentech, the RHHBY subsidiary (and dominant entity), has 14 marketed biooncology products and another 25 candidates in its pipeline. It has the structural, longer term advantage over all the other contenders. It can study, for example, Tecentriq with its own angiogenesis inhibitor Avastin. Since the PD-1's work in lymphoma, I suppose it could study Tecentriq with its Gazyva (maybe it's doing so). And so on. There are competitive, financial and scientific benefits from doing the combos fully in-house. I actually bought a small amount of RHHBY for the dividend - though I don't even know if it's fully secure. I'm now broadly exposed to biotech qua biotech (not biopharma). Dull, sort of like owning McDonald's (NYSE:MCD). But presumably resilient and dominant for many years.

Granted that it's not really proper to kick a company when it's down, but it appears fair and accurate to say that BMY put all its NSCLC eggs in one PD-L1 basket and got burned. We see the result with a 1/3 market cap loss. This raises the question of why BMY did not also do a Keytruda-type study. How could it have been so certain the hypothesis would pan out? As shown above with the Genentech drug, these stories are new, complex - and unpredictable. And now, for lack of either a separate study using precision medicine as MRK did with Keytruda, or a separate 50% arm in the above failed study, BMY is suddenly behind the 8-ball in one of the most important therapeutic categories in the world.

At the same time, the Keytruda study is now a target for the next guys to go after. As with Gilead's (NASDAQ:GILD) lead in HCV, the other guys can catch up sooner than one might think. So I just don't know how to value the future cash flows from Keytruda. And the rest of MRK is... what it is. Richly valued in my opinion.

In contrast, unique products that are years ahead of the competition, such as REGN's dupilumab for allergic disorders, attract me more. I've spilled enough digital ink on REGN, though, to move on to make some points on the market (NYSEARCA:SPY) at large based on the above observations.

BMY's crash and growth stocks

At a recent price of $75, BMY was trading at about 30X projected 2016 GAAP EPS with a dividend yield of about 1.7%. Therefore, a growth stock.

For a large company with many non-growing parts to trade at that valuation, the stock market has to be dangerous, even frothy.

As Seth Klarman might say, and Ben Graham before him, "dude, where's my margin of error?"

It's really not appropriate that one failed clinical trial could do this to shares of a powerhouse major company. This is the sort of crash that one might expect if Opdivo were suddenly pulled from the market, rather than not being able to gain one indication.

These crashes continue in the sector. Monday after hours, look what happened to another, even higher-P/E biotech growth stock, Illumina (NASDAQ:ILMN) - a huge crash of 26% on some revenue weakness. Yet the future prospects may not have changed much.

I won't discuss ILMN, but the general concept is similar; an attractive market has a margin of error for the bad stuff.

Per the WSJ, the P/E on the NASDAQ 100 is 24.4 as of 10/7, up from an already healthy 20.75 one year earlier. The SPY, with weaker growth prospects than the NASDAQ, trades at 24.6X as of 10/10.

Probably even more dangerously, the Dow Utilities (^DJU) trade at a stated P/E of 26.4, something I'll come back to in a minute.

The reason for going through this analysis now is that, to come back to BMY, no one really knows how to value BMY's 2-year, 5-year, 10-year, and 20-year growth rates. One has to think this far out given a 20X P/E as of now based on 2016 expected EPS. Maybe the "right" P/E should be 18X, where Novo Nordisk (NYSE:NVO) trades? That would imply a further 10% loss. But maybe interest rates will rise, Hillary will win a wave election, and pharma stocks will run into further significant headwinds? So maybe BMY in reality "should" have a 14X P/E, for a further 30% downside risk? Or at least could trade there?

In 2011, BMY earned $2.16/share per Value Line (I'm not sure if that's GAAP or non-GAAP) and traded as low as $25 both at the beginning of the year and after the summer sell-off. Things have gone very well for the company since then, but as GILD shareholders know very well, they ring no bells at or near the top. Sometimes prolonged and severe P/E compression occurs, and sometimes the "E" does not meet expectations.

Maybe Opdivo will turn out to be an also-ran, as it were a Pravachol (a BMY blockbuster) passed first by Zocor, then Lipitor. And maybe the vaunted pipeline isn't so hot.

All this could well apply, one way or another, to the valuations of "growth" stocks all over this market.

This sort of thinking is what leads to bear markets when it applies to the market as a whole even without a recession.

That's BMY as a growth stock, now at a P/E above traditional value stocks but below that of the general SPY, thus not valued as a growth stock anymore.

A comment on my prior articles on BMY:

I first reviewed BMY after it reported Q1 earnings this year. Based solely on valuation, not expecting this poor result for Opdivo in the -026 study, with BMY around $72, I set around $62 as my buy point, which was 25X estimated 2016 EPS. That was the valuation I paid for NVO and as high as I wanted to go. However, when BMY got to the low $60s, it was only with the announcement of the negative top-line results from -026. I reviewed it and suggested that $50-55 was probably where I thought the stock was now attractive. Of course, by then, NVO and GILD had been collapsing. The hope in August is that though the top-line results were poor, subgroup analysis could well be positive, showing similar results (it was hoped) to the Keytruda data, for which only topline data had been released.

Now that we are "here" at the $50 range, I'd probably add a little BMY if I were a portfolio manager with excess cash. But as a small investor, I'm not really loving the vibes emanating from the fallout of this failed study. So I'll just watch it. With Hillary the clear favorite now to take office, there's no need to do more than just enjoy RHHBY's dividend stream as a small new position and possibly multi-decade leading position in the biotech sector. It's far safer than BMY's.

So, BMY as a growth stock looks to me paradigmatic of how risky it is at today's valuations.

In a recession, how low do you think (NASDAQ:AMZN) can trade? Netflix (NASDAQ:NFLX)? Tesla (NASDAQ:TSLA)? If investors want "out" of those names, will they really care if the 10-year Treasury is at 1% or 3%?

No. There was never much of a relation between those asset classes, at least not a predictable or consistent one. To assume this correlation will continue is a gamble in my view.

BMY and income stocks

Value Line shows BMY averaging a 5.0% dividend yield in 2011. It averaged 4.6% in 2006, when it did not even earn its dividend.

As of Monday's close BMY's dividend yield was 2.74%. Is that attractive? Not really. A general bear market decline of 20-25% could very easily produce a 10-15% further decline in BMY, even if it is by then treated as a safe haven.

To be truly an income stock, and assuming even modest reversion upward toward the mean in interest rates, a stock has to yield 4%. Actually, in the golden era for stocks in the 1950s into the 60s, the dictum was to sell the market (NYSEARCA:DIA) if dividend yields dropped to 3% and buy it when they rose to 6%. Meaning, 4% really can just be considered normal, not really "income stock" level.

All this obviously ties into the current era of low to vanishingly low interest rates - but those can change. They ring no bells at the bond market top, either.

To go from a 2.7% yield to a 4% yield quickly, with no dividend increase, requires a 1/3 drop in price, implying a new price of $33/share.

No one is yet thinking of BMY as a true income stock, but now think of electric and gas utilities (NYSEARCA:XLU). Per that WSJ link, the Dow utilities (not the same as the XLU stocks) yield 3.47% and that amazing 26.4X P/E. Undoubtedly there are "special items," but the XLU website lists a P/E of 21X. Projected 3-5 year growth rates are 5%.

In a US growing at a nominal GDP of 3-4%, this is an exaggerated P/E. Given conservation measures such as ever-increasing efficiency of appliances, this 5% projected EPS growth rate is optimistic.

Utilities and many other stocks that trade at high P/Es and dividend yields that are competitive only due to low competing bond yields have two risks, whereas bonds only have one of those risks. These equities are operating companies. If the bond market turns out to be more right than wrong, i.e. not in anything like a "bubble," then forward growth could "go Japanese" and turn toward zero nominal revenue growth, not 5% or whatever the projections are. Higher quality bonds would do just fine in that environment, though.

The opposite, a rising inflation/rising rate scenario, can easily be worse for "income stocks" such as utilities than for an individual bond. The bond matures and ends; one's money comes back. So the investor loses something to inflation. But a stock is perpetual. The dividend may or may not be able to be raised, but a typical finding in times of rising inflation is that regulators do not allow utilities generous rate increases, so dividend increases are slow. The upshot can be years of negative total returns as, say, Mr. Market eventually wants his 7% yield of old on the XLU, but dividend increases do not come close to letting it get to that yield without sharp stock price declines.

BMY as a dividend stock in 2006 and again in 2010-11 had the advantage of interest rates that were about to decline. Shareholders may not be so lucky this time, which could apply to the SPY as a whole.

Now to sum up and put all the threads together.

Biotech, immuno-oncology and the market as a whole

One of the really great things about the US economy and markets is how capital creation can produce a multiplicity of the next New Thing. In this case it's PD-1 antagonists for cancer. But the attitude of investors changes greatly over the years. Now we have optimism embedded in stock valuations. In prior eras, BMY would not have gotten near 30X forward EPS even with low interest rates. Investors "in the day" wanted an old-line company such as BMY to actually deliver, not bet on the unknowable of how a clinical trial would come out and then how much competition would actually come in to limit returns.

It's that old-line conservative or fearful mindset that really rewarded the brave, and in 1948 or 1953, it took some bravery to commit big-time to stocks, with the memory of the Depression and war fresh in every investor's mind.

Now, the zeitgeist is that either the economy will deliver, or central banks will bail out stock market investors.

That works until, one day, it doesn't. And just as BMY has lost 1/3 of its value quickly, with no change in its operating metrics, the SPY may do the same - for any reason, or no obvious reason at all.

More broadly, Tesla could not have been supported at or near current valuations. The investor base demanded safety of principal, as with a bond, plus income from dividends greater than a bond. That was in the 1940s and 50s.

Eventually that gave way to requiring high yields on both stocks and bonds.

Now we are in a period that I believe is the single riskiest in US history for investors, at least going back to the immediate post-Civil War era. Just as rents are too d--n high, yields are too d--n low. I saw a headline yesterday that 70% of Americans have almost no savings - less than $1000, if I remember correctly. So the Fed has "printed" all this money, but to a large extent, it has just been cycling endlessly and with increasing recklessness amongst the monied set.

Even in the short term, risks abound. The Chinese currency is just today (Tuesday) breaking to new multi-year lows against the USD. This has been associated with some market breaks in the past. Jeff Gundlach notes that one of his recession warning triggers, a "golden" (or bearish) cross of the unemployment rate against a moving average, has now occurred. But he sees no recession. The forecasters ECRI and Conference Board, which computes the LEI, see no recession. But:

  • it's almost 9 years since the last recession began
  • valuations are so high, a bear market can occur just on P/E compression alone
  • recessions can come on quickly.

To finish with two almost opposite points:

While central banks have been the conductors of the risk-on bands in the financial markets, investors have played most of the actual instruments. Very high and in some case record stock valuations have joined, for the first time I know of, record low interest rates up and down the yield curve and up and down the credit range. This is so antithetical to the attractive, depressed markets and economy of the 1979-82 or 1984 period that I continue to believe that as one should have avoided the siren song of high short-term rates then and embraced duration, it may well pay to do the opposite now and underweight duration. Since equities have traded up in valuation with bonds over the years and have also traded up with the economic cycle since 2009, and since each share of stock is perpetual, I'm inclined to view them as the riskiest asset class. This view is supported, for example, by the rampant speculative passion for stocks seen on the US futures exchange for the NASDAQ 100. Per FINVIZ:

With the Fed not increasing the money supply for almost 2 full years, I don't know where the fuel for higher valuations is coming from. Maybe, but as you see on this multi-year chart, the large speculators in the mini contract (bottom panel, red line) are heavily bullish. That's at the all-time high in price. That positioning only worked in QE 3, 2013-4, as far as this chart goes. But in 2012 and periods of 2015 and again this year, this degree of trend-following optimism has been negatively rewarded with sharp sell-offs.

More broadly, my thesis that the end of QE 3 would lead to an environment favoring bonds over stocks has panned out well so far. QE 3 ended at the end of October 2014. The SPY closed October 2014 at $202; it's now $215, an increase of about 6.5% per year. That's only been accomplished due to P/E expansion, as earnings of the SPY have actually shrunk since then (not in any mainstream projections). A popular long T-bond ETF (NYSEARCA:TLT) has increased in price from $123 to $133, a somewhat large amount with higher yield. Importantly, TLT has had lower duration and "should" have underperformed SPY in this risk-on environment.

A fairer comparison with SPY is a zero coupon T-bond fund (NYSEARCA:ZROZ), where the duration is more like that of SPY. Even though SPY was in a sell-off exactly 2 years ago, ZROZ has had an identical price return going back to Oct. 11, 2014, but with higher yield.

I take the failure of the most publicized, hyped market, the SPY, to beat the un-hyped bond market of October 2014 is a warning sign.

And, both TLT and ZROZ are solidly beating SPY this calendar year, with greater theoretical safety. To me that suggests that a reasonable, evidence-based approach continues to be that the farther in time we in the US get from full-on QE by the Fed (only the Fed can create USD).

On the other hand, a different message comes from looking back at BMY or, this year, one of my permanent holdings, Apple (NASDAQ:AAPL). These show how valuation matters. The BMY of 5 years ago embodied the potential for Eliquis and I/O assets to become stars. If not, BMY probably always had the option of selling itself, or parts of itself, and rewarding investors thusly.

Today, we see that with AAPL back up to $117, having traded for a while as an unloved value stock, patience has been rewarded. With the stock in the $90s, shareholders got a similar deal as I've been pointing out that GILD offers right now. Ex-cash, AAPL was offering a low double-digit earnings yield. Meanwhile, the basic business rolled on. Since maybe half of life really is just showing up and doing your job competently, just as AAPL is benefiting from the unforced Samsung (OTC:SSNLF) error, there are some good things that are not in the GILD share price, at below 7X TTM EPS.

One such good possibility for GILD is that ViiV may fade. Its 2-drug combo for HIV may not look impressive when the trial results are in. If so, GILD with TAF and bictegravir, plus the old reliable emtricitabine, may along with its other drugs move to a much higher market share. If so, almost all of that market share will be pre-tax profit. So while my forward guesstimates for GILD's HIV/AIDS line may be too high, they also may be too low. That's the AAPL scenario right now.

In other words, the flip side of the opinion I'm presenting is not to think too hard when an apparent bargain in a high quality asset appears.

A final thought. Outside of short-selling into a bull move, one of the single hardest thing for investors to do is sit out a bull move. But if the numbers do not work for me, I've learned to do it. Similarly, another extremely difficult thing for investors to do is wait out a sharp bear move in what appeared to be, and basically still appears to be, a good asset. Think AAPL both recently and in 2012, or Treasuries into the vicious bear market in the Taper Tantrum of 2013. Or GILD today.

There are no guarantees, and buying the dip tends to be a source of short-term frustration. But then the next year arrives, and the year after that, and with them possibly entirely new prices. High P/Es must be justified over time with substantial and durable growth, whereas low P/Es have a much lower bar over which to jump to make the owner of the stock happy.

With I/O start-ups all over the place, extensive investment in the sector widespread now in the biotech industry, I see this sector in general as richly valued and not having special investment interest even to myself. To the extent that I have it covered with some exposure to RHHBY and REGN, I feel more than exposed enough.

As far as stocks go, an investment buddy and I were emailing a month or two ago. He opined, quite correctly for now, that $215 on the SPY was an important support level. Without disagreeing, since I have almost no interest in the short term these days, I responded that I was more concerned not about $215 but about $125 - one of these years. That would be roughly the price the SPY would be near if it fell as much from its all-time high as much as BMY fell from its all-time high. If trailing EPS on the SPY is around $100 after Q3 or Q4 are over (GAAP preferred until a major write-off period occurs), then a $125 SPY would simply be a 12.5X P/E. The market has spent a lot of time at and below, sometimes far below, that P/E going back to 1871 (link is to an interesting table). I believe it can happen again.

Whether it will do so is for the future to show us, but all we as individual investors or managers of other people's money can do is play the odds as we see them, and sleep well at night within our risk tolerances.

I see the BMY example as instructive. Shares of big, strong companies simply do not collapse by 1/3 or more just because one individual clinical trial, of the many ongoing and planned, goes poorly. Yet here we are, and at 20X expected earnings, it cannot be said that BMY is even now cheap.

The same is true of "growth" stocks in today's market.

Unfortunately, on the opposite side of the ledger, income stocks have, within their expected risk range, almost equally grim possible results. Rising inflation, rising rates; or recession/Japan scenario: each can do real damage over many years to patient investors in the XLU and soap and cola stocks that also trade as bond substitutes. Whereas, the individual corporate bonds simply have to not default to deliver a positive total return.


I see risk everywhere now in the US financial markets. I do not expect the Fed to bail me out the way it bailed bankers out. I continue to see high-yielding insured bank deposits, higher-yielding money market funds and near-cash investments such as short-term fixed income funds as reasonably attractive vehicles to provide small amounts of income with no, or almost no, risk to principal. I believe that right now, the "don't fight the Fed" mantra joins with the fact of low interest rates and low earnings yields in the stock market to create strong headwinds for patient investors who are long duration, especially in US-listed equities.

The great advances in immuno-oncology, requiring very large sums of risk capital, paradoxically provide an example of how an economy or a sector of the economy can function well, but investors may find it hard to make a buck.

Thanks for reading and for any observations you might wish to share.

Disclosure: I am/we are long RHHBY,GILD,NVO,TLT,ZROZ.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: Not investment advice. I am not an investment adviser.

Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.