Bronte Amalthea Fund commentary for the month ended December 31, 2016.
In December the fund significantly outperformed strong global markets in both absolute and relative terms. Our longs generated the gains and performed better than the market whilst our substantial short position was neutral which is unusually favourable in such a strong month.
This being the end of the quarter we provide a lengthier commentary in two sections. In the first we present a detailed discussion of a new long added in the quarter, which is Bayer. Second, in something of a departure for us, we offer some macro commentary that explains the positioning we have built from our micro-level research.
Bronte Amalthea Fund – Bayer
We have added one new big position to the portfolio in the last quarter. We now hold a 5 percent position in Bayer – the largest company in the German index. Bayer has four and half businesses which they collectively call “life sciences” but in fact fit into no specialist portfolio.
These businesses are:
(a) A large second tier pharmaceutical company which has a few major drugs, on which they have partnered,
(b) A large over-the-counter (OTC) medical products business – which they (plausibly) argue is the biggest in the world,
(c) A moderately large seed technology business
(d) A moderately large animal health business.
The half business is a majority stake in Covestro – Bayer’s old materials business.
The whole contraption trades for between 2 and 2.5 times revenue depending on how you net out Covestra. This is really cheap for this type of businesses. Ultimately most of these are pretty good businesses that earn high margins, should earn even higher margins and normally trade at bigger multiples.
The main knock on the stock is that they are buying Monsanto to create by far the biggest seed technology business in the world. They are paying well over 4 times revenue for Monsanto. It is far from obvious financially why they should do this acquisition.
Moreover, it is pretty clear that the market hates this deal. This is now too big a seed technology business to attract pharmaceutical focused investors. Monsanto is a company that sells genetically modified seeds (unpopular in Europe) and pesticides (also unpopular) and sues farmers for not paying seed royalties (unpopular everywhere).
As far as the market is concerned this is a big bad deal.
The businesses and comparisons
Bayer is a genuinely large pharmaceutical company. The right valuation comparison is the mega-giants. Pfizer trades at 4.2 times sales. Novartis trades at 4.1 times sales. We could go on. Most pharmaceutical companies trade above 4 times sales.
Bayer has big advantages over many of these companies – most of its key patents have a lot of years left to run, and it has a decent pipeline. The drugs have obvious revenue growth.
The biggest problem is that the two main drugs have issues, one essentially legal, and the other potential low cost competition.
The two biggest drugs are Xarelto and rights to Eylea which they share with other companies.
Xarelto is a Warfarin substitute. Warfarin is an old drug – out of patent and cheap. It is a blood thinner – and reduces stroke risk and other risks particularly in old people. It is very widely prescribed (including maybe 10 percent of the population of the Western world over the age of 80). And it causes considerable problems.
Dosing is tricky. The difference between an effective dose and a dangerous dose is small and variable. If you give someone too large a dose they can get internal bleeding (especially internal bleeding in the brain) and die. In fact it is pretty good at causing death, and so Warfarin is also used as rat poison.
Warfarin has one big advantage as a blood thinner and as a rat poison. There is a cheap, easily available antidote – just a big injection of Vitamin K along with some plasma.
Xarelto has big advantages over Warfarin. Most importantly the dosing is easier. It is just easier for a general practitioner to get dosage right and hence the doctor is less likely to kill the patient. This is an enormous selling point and makes Xarelto a drug of choice outside hospital settings.
Secondly it is claimed that Xarelto preferentially causes bleeding in the intestine (where it is passed and noticed) rather than in the brain (where it unnoticed and very dangerous).
There is one big disadvantage over Warfarin: there is no antidote.
The big problem with this drug though is that it is a blood thinner – it will cause internal bleeding, and that will cause death and/or brain damage. If you sell these things you must expect considerable litigation costs even if, on average, they do considerable good.
The second important drug is Eylea – a treatment for wet age-related macular degeneration.
Wet AMD is a leading cause of blindness in old people – and stopping people going blind is a good formula for getting very high prices for your drug. If your drug stops people going blind it has huge economic benefits and medical systems (even government funded ones like Australia) will pay a lot for this.
Now let’s get past the squeamish bit. This works by monthly injections in the eye. That treatment regime means that non-compliance is high (even though noncompliance leads to blindness).
But alas the story gets more complex. There is a relatively old cancer treatment drug that stops the body making more capillaries. The drug is Avastin and it is marketed by Genentech (a subsidiary of Roche). It is a very important cancer drug – the idea being that you zap the tumor with radiotherapy or similar and then give Avastin to stop new capillary formation and hence the regrowth of the tumor.
If you inject Avastin in very small quantities in the eye on a monthly cycle you will slow the development of Wet AMD.
The problem from the drug company perspective is that very small doses of Avastin are cheap (maybe $50). And this is a treatment the drug company really believes they should be paid for.
So Genentech came up with a very similar drug (Lucentis) which works the same way as Avastin and which they tested as a treatment for Wet AMD. It works.
They charge a large amount for it (about $24,000 a year or $2,000 an injection).
They never subjected Avastin to a proper double-blind study. As Avastin is not “proved” to be effective it is not prescribed – and so Genentech gets lots of extra money from selling Lucentis instead.
Eylea is yet another version of the drug developed by Regeneron (listed in the US) and Bayer. It also gets a lot of money and is priced about the same as Lucentis.
Whether these prices are sustainable is open for debate. Genentech are hardly going to offer up their cheaper drug by subjecting it to an expensive double-blind trial. When we looked at the treatment regime subsidized by the Australian government we found only Lucentis and Eylea. Avastin was not allowed. Concerningly however, some European governments dictate Avastin.
Again this is a really profitable business – we are not entirely sure how sustainable. But the Australian government is far more rational about drug pricing than most – and it is not doing the cheap thing here.
We think that Bayer’s pharma business is not worth much of a discount (if any) to some of the majors despite these problems.
The OTC medicines business
Bayer has what is probably the largest OTC medicines business in the world. This is inherently very fat margin. The only brand that everyone seems to know is Asprin. They developed the product and still own the brand name.
Some however are designed for super-fat margins. For example Bayer market Elevit – probably the best-known pregnancy vitamin supplement including folate which is known to stop some major birth defects. If you want the best fat-margin pitch you can think of, it is telling a middleclass woman that if she takes this her baby will be smart and healthy.
The right comparable for this business is Reckitt Benkeiser who market consumer products and OTC medicines. RB trades at over five times sales. It has a slightly fatter EBITDA margin but quite a lot fatter EBIT margin. We think (with good reason) the margin of Bayer’s business should increase a fair bit from here.
This part of Bayer is clearly worth well over 2.5 times sales.
The seed technology business
Seed technology is inherently a wonderful business. It is a very small thing (say a seed coating) that makes a big thing (the year’s output for the farm) better. If it is demonstrated to work once the farmer is reluctant to experiment and try other things (it really matters to them). And it is a consumable – the farmer is back every year.
This is right-up Bronte’s alley and we have interests in other seed companies as well. And the good bits are clearly worth well over 2.5 times sales.
There are bad bits in this. Bayer has a neonicotinoids business. These are insecticides and they are very controversial – the allegation being that they are the underlying cause of global problems in bee hives. Monsanto also has a neonicotinoids business.
We think this bit is likely to be subject to regulatory pressure and is the main part of the business we think is worth less than two times sales.
The animal health business
Some of this is really profitable – e.g. pharmaceuticals for your pets. Some is more competitive like pharmacy for livestock, but it is really quite a good business. The most important competitor (Zoetis) trades at six times sales – and this business is of similar quality. We hope however they sell to Zoetis because they will get a good price. We are unsure of the future here.
The residual materials business will be sold.
The observation here is that apart from smaller things like the neonicotinoids business there is very little here that you would not pay 3x sales for (especially in this market) and there are several things you would pay 4x or more times sales.
The whole thing is much cheaper than that and we think there are very few ways to lose here.
The big problem is that Bayer is buying Monsanto at a much bigger multiple than itself.
For the average American CEO the solution would be obvious: do not buy Monsanto, buy your own stock back. This is Germany and that is not going to happen. German supervisory boards have union representatives – and as the company’s solvency guarantees pensions the boards are opposed to buy-backs.
So the company is doing a highly priced and unpopular acquisition.
There is a joke here: buybacks would increase stock price and you should never stand between an American CEO and a big pile of money. Buying Monsanto is a path to global domination – and you should never stand between a German CEO and global domination.
The joke is not entirely accurate – but there is more than a seed of truth.
Ways to win
We think this is one where it is very hard to lose. If the Monsanto deal closes the under-utilised balance sheet of Bayer will be levered up fairly hard. The price-earnings ratio of the stock comes down – probably to well under 12 and maybe as low as 10. The stock will go up if this happens.
However economic nationalism has raised its head. There is a real possibility, which American journalists put at 50-50, that the Monsanto deal will be politically blocked. The stock will also go up if this happens.
This is not the highest quality company in our portfolio – but we have not had to move the quality standards very much. It is certainly a way higher quality business than average.
And we are not paying a lot for it.
The Big Picture
We are not a macro-fund and we do not pretend to macroeconomics. There are a few macro-players we admire – and who we read. But our day-to-day is in the minutia of businesses which we are interested in – like Bayer above.
So our perception of the market overall is usually distilled from our bottom-up research. We believe markets are expensive when the good companies we find cannot be bought at prices we consider attractive and when we find a plenitude of iffy companies at high prices to short.
Conversely, we judge markets to be cheap when we find lots of bargains on the long side and our beloved stock promotion shorts have fulfilled their destiny by closing asymptotically on zero.
From our perch, we think markets are expensive, particularly in America.
As a result we are running as close to neutral as prudent risk management allows. We are – on a beta-adjusted basis – net short in America and long in Europe.
As experts on individual trees, it behooves us to occasionally have a look at the forest in which they’re growing. That is, to understand the macro environment in which we’re operating, even if we do not attempt to specifically forecast it ourselves.
Debating the Macro Environment
One recent round of vigorous internal discussion—which we seek to encourage—questioned this bottoms-up perception of the US market as being expensive. The price earnings ratio of the S&P 500, the counter-argument goes, is not particularly high by historical standards, therefore there must be plenty out there to buy (at least on average). So we tried to quantify just how expensive the market is from a top-down perspective. When the downturn comes (and it will) this is going to be a guide for when we should get more long rather than be neutral.
S&P Earnings for the last sixteen quarters
Below are S&P 500 earnings by quarter since when the dotcom boom really took off:
The source is Professor Aswath Damadaran and the last year is an estimate (as the data is not in yet). You will notice that S&P 500 earnings have stagnated in the last few years. Almost all the headwind was the result of falling oil prices and the consequent drop in earnings for oil companies, especially the majors. There has been some offsetting rise in profits in the consumer sector and if you think those are related you are probably right. Consumer spending (on things other than oil) and hence consumer profits are up with the oil price decline.
On January 10, 2017 the S&P 500 stands at 2269, with a trailing 12-month price-to-earnings (PE) ratio of just under 21. This PE ratio is elevated versus history but not astoundingly high especially if compared to the peak of the dot-com bubble when it nearly touched 30. Moreover, interest rates are much lower than they were then, so PE ratios should be higher now than at most times in history. Taking these numbers seriously means the S&P looks high but not astoundingly so. If anything, the index is 10-20 percent too high, which is high enough to make one cautious about future return expectations but not so high as to imply large and imminent permanent capital losses.
The focal point of our debate thus became, “Is this view too bearish?”
Perhaps we aren’t finding value so readily, not because it has become an endangered species but because of a fault in our process. Though we will always believe there is room for improvement in our research process—this is a business of continuous learning, after all—we take you through the evidence that we believe supports our view.
1.The price earnings ratio in the dot-com period was illusory.
In the dot-com era there were a very few stocks (some of which you have likely never heard and many of which are consigned to the dustbin of history) that had enormous market caps and little or no earnings. For instance, i2 Technologies at one point had a market value over $200 billion. If one excludes from the overall PE ratio the most absurd bubble stocks, the PE ratio in (say) the year 2000 was far more modest.
Professor Ken French (of Dartmouth) compiles a lot of data which we find useful. One table works out the price earnings ratio of the median stock on the New York Stock Exchange.
The chart isn’t quite updated to the present. The current PE ratio of the median stock is an all-time high.
We do not invest in “average stocks”. We invest in the best stocks we can find (which we hope of course to be way better than average stocks). When the median stock is at an all-time high in price earnings ratio our pickings are likely to be thinner than average.
2. Earnings are going to fall
The table above showing earnings for the S&P since 2008 shows a very good upward progression. Sure, earnings fell during the crisis as banks destroyed a lot of capital. They even fell during the 2001 dot-com bust. And the latest oil price swoon has lowered them as well. Despite that, the past 20 years or so have been a golden age of corporate earnings growth. Indeed, US corporate profits as a percent of GDP have returned to all-time highs last reached just before the Global Financial Crisis and the “Go-Go” era of the 1960s.
The idea that earnings may not be higher in five years is very anti-consensus.
We have no idea where they will be in five years, but Bronte is fairly comfortable that S&P aggregate earnings will be down in twelve months. (Again this sort of macroprognosis is not our usual bailiwick – so you can expect that there remains debate within the firm and that is a normal part of the process. Also, given the non-GAAP games we see companies playing lately, “earnings” have proven to be a malleable concept.)
The reasons earnings have been up so much are well known. The S&P has internationalized and large US companies are catching a larger share of a global pie. Economic growth has been okay and wage growth has been slower than GDP growth for a fair while.
The low level of wage growth is obvious to everyone now (and is reflected in politics). Say what you will about low wage growth – it has been great for profit share.
We would argue that both of these trends—internationalization and wage stagnation— are weakening. A protectionist/anti-globalist streak has appeared in politics. Moreover the US dollar has
strengthened considerably, and mere translation diminishes the US dollar value of foreign earnings.3
But more importantly US wage pressure is clearly upward. Anecdotal evidence abounds, such as Walmart voluntarily giving pay rises to their staff (because they had trouble holding on to them). But the best aggregate data comes from the Federal Reserve Bank of Atlanta. Their wage growth tracker chart shows:
For the first time in a while, nominal wage growth exceeds nominal GDP growth. The sub-indices in this series (such as wage growth for people who are shifting jobs) suggest accelerating growth in wages. This is, of course, great for workers. But it is not so good for S&P earnings.
Jonathan Tepper of the London research firm Variant Perception, a friend of Bronte, also provided us this seemingly unlikely chart. Below we show, over the last 60 years, periods when unemployment in the US has been low (defined as official rates below 5%) graphed against the subsequent five-year returns on the S&P 500.
This, for many, may be a deeply counter-intuitive result. When the economy is performing well, the subsequent five-year returns of the S&P 500 are mostly negative, and low always.
There are good reasons for this. Firstly, when unemployment is below five percent (currently 4.7 percent) the economy has done well and valuations are high. This is certainly true today.
But more importantly labor earns a large share of GDP (typically over 50 percent) and the corporate profit share of GDP is low (typically around or below ten percent). Small changes in wages have a levered effect on profits. When unemployment is very low, wages grow faster than GDP and profit shares shrink. Again this is supportive of earnings weakness.
So in aggregate the S&P median price earnings ratio is at record highs just as S&P earnings look like they are going to fall.
While either one of these factors might cause a 10-20 percent decline in prices, combined they could imply the risk of a decline closer to 20-40 percent.
The Trump Offset or: How I Learned to Stop Worrying And Love the Tweets
The market is euphoric about the Trump administration. But our argument for earnings weakness above is policy independent. We think that pre-tax earnings for the S&P fall regardless of Trump’s policies.
But post-tax earnings are within President Trump’s power to change. Large tax cuts could increase post-tax earnings per share. This is a one-off increase and will to some extent be offset for valuation purposes by higher interest rates (induced by looser fiscal policy). Tax cuts give the market one last reason to celebrate.
(Note that the credit markets had quite the opposite reaction to Trump’s victory as the equity markets. A tighter credit regime will hurt non-financial companies’ earnings and perhaps diminish lenders’ willingness to be as accommodating to semi-distressed companies as they have recently.)
The Buffett measure
Warren Buffett has used a measure of the aggregate level of the stock market that appeals to us. The measure is stock market capitalization as a percentage of US GDP. The reason this appeals to us is that it looks through the fact that margins/earnings are cyclical. Here we are using a chart from the St Louis Fed showing we are at all time highs and very large falls are possible.
It is however fair to say that this measure should trend up over time. US businesses are capturing a larger proportion of global business than, say, in 1970. Further, the big companies now (such as Google) are probably less subject to competition than big companies of 50 years ago (such as US Steel). Because they are less subject to competition, the earnings as a proportion of GDP should be higher now than 50 years ago.
That said, this measure is supportive of the idea that markets are more expensive than the simple PE ratio of 21 would suggest.
As historically-minded readers will no doubt remember, Buffett somewhat infamously spoke at the Allen & Company conference in Sun Valley in July 1999 to an audience of true believers in tech and media (a group of people who had gained much from the bubble’s inflation). Buffett challenged his listeners to think about the future direction of the trifecta of macroeconomic factors supporting stock valuations: earnings share of GDP, interest rates, and tax take. Earnings share has grown tremendously, interest rates have hit all time lows, and tax take is only a little lower. But where will we go from here? US interest rates are rising and corporate earnings are under pressure, though tax take may decline. This does not a case for unbridled optimism make, in our view.
What Bronte is Doing
Rest assured that we have not given up the hunt for value in North America. While bargains are hide to find, it does not absolve us from trying. And we do.
But if we don’t add very much to the North American net exposure, it is because we do not believe one should lower one’s standards for quality (much) when markets become expensive. That said, we continue to look for high quality businesses we can own at reasonable prices in the US and globally.
There are two iconic ways value managers get themselves into trouble. The first John blogged about recently: by “averaging down”—buying more as the stock drops— when they are wrong.
The second is by purchasing lower and lower quality stocks as the market price increases. When stocks are cheap, good quality companies with great prospects will sell at prices less than ten times their normalized earnings.
But when markets unerringly rise, the only stocks that look so cheap (and which attract value investors because they screen well) are cheap for a reason. Many a fund manager has come undone by lowering standards in a bull market. We will not do this.
Value in Europe
We have added a few decent sized positions to the portfolio over the past year – notably Bayer (from Germany) and Elementis (which is listed in the UK but is really a US dollar company). Both of these were at reasonable if not spectacular values.
This has left us a currency imbalance which we have fought for a while. We are overweight Europe and as we did not have much currency hedging we wound up being overweight the Euro.
If you look at our returns over the past twelve months we have been way better than average stock pickers (okay on the longs, okay on the shorts but without spectacular wins). But we have mucked up currency badly (being underweight the very strong US dollar).
As we are not macro-economists we don’t pretend to expertise on currency. Sometimes we will get it right, sometimes wrong, but we will not swing for the fences on it ever.
If our stock performance stays where it is and we stop getting currency wrong our returns will be consistent with our goals. If the US markets goes down especially relative to European markets we could do very well indeed.
And we have a few things in the portfolio (most of which we do not wish to go public on) that might make us truly outsized returns. Most of those have not worked recently but they have worked in the past and they should work again sometime. We look forward to it.
Thanks again for the trust you place in us.
The Bronte team
Post script: this letter and the data that it contains was compiled with the help of Jonathan Tepper of the London research firm Variant Perception. John likes him and recommends him strongly.