I’ve sold AstraZeneca because of its rising share price and weak results

When I added AstraZeneca to my portfolio in mid-2015 it was the second time I’d invested in this well-known pharmaceutical company. The reasons for both investments were more or less the same:

1. AstraZeneca appeared to be a fundamentally sound company with a good track record and a strong balance sheet.
2. The share price appeared to be attractive because of the uncertainty caused by its “patent cliff” (where major blockbuster medicines see their profit margins collapse as patents expire).

The investment basically came down to a few assumptions which were directly related to those two reasons:

1. AstraZeneca had a long history of profitable medicine development, so new medicines might be developed quickly enough to replace the old.
2. The company had a strong balance sheet, so it could use debt to fund research and development and/or the dividend during the lean years.
3. At some point, investors would probably become more optimistic about Astra’s future and the share price would go up.

And that’s more or less what happened, although the final results were satisfactory rather than spectacular.

Those are the final results in dry numbers and an 11% annualised return is satisfactory because that’s slightly above my target return of at least 10% per year over the long.

But the journey to that final outcome was relatively volatile, especially from a large-cap “defensive” sector company:

So that’s the big picture. In the rest of this post-sale review, I’ll go over why I bought AstraZeneca, why I held on to it and why I’ve sold it in a bit more detail.

Buying: AstraZeneca was a defensive sector company facing serious headwinds

Pharmaceutical companies are classified as defensive because demand for medicines is only slightly impacted by the economic cycle, but in reality (and as the share price chart above shows) they’re often anything but defensive.

However, unlike most cyclical businesses, their cyclicality comes from the lifecycle of individual medicines rather than the economic cycle, and for AstraZeneca, the medicine lifecycle is measured in decades rather than years.

For example, it can take more than a decade just to get a new medicine to the point where it can be tested on humans.

If it passes all the required regulatory tests (which can also take years) it can then take another decade to increase the medicine’s sales from zero to their full potential.

After that, there might be another decade of sales which are highly profitable thanks to the protection of the medicine’s patents.

Patents exclude other companies from producing too-similar products, but those patents don’t last forever. After 20 years or so the patent will expire and “generic” drug companies will enter the market with copy-cat products at far lower prices (just compare the cost of Reckitt Benckiser’s Nurofen to generic Ibuprofen).

This inevitable decline towards lower sales and thinner profit margins marks the final chapter of a medicine’s lifecycle.

From a financial point of view, there are two main phases to this lifecycle:

• The first involves huge amounts of spending on R&D as ideas for new medicines are researched and developed. Most of these fail, and no revenues or profits are generated.
• The second phase involves the sale of patented medicines, and this phase involves little in the way of expenses and much in the way of revenues and profits.

When I bought AstraZeneca in 2015 it was at the tail-end of the second phase, with major products going off-patent, including Crestor, Nexium and Symbicort, which together were responsible for almost 50% of the company’s revenues.

That’s a huge hole to fill with new products, the development and commercialisation of which were (and still are) largely uncertain. To say the company was facing an uphill struggle would be an understatement.

However, I thought the company stood a good chance of at least treading water over the next few years thanks to its vast research and development capabilities. And with a dividend yield of 4.4% at the time, I thought treading water and maintaining the dividend would be good enough to produce decent returns, especially if Mr Market became optimistic at some point.

That may not sound like a ringing endorsement of AstraZeneca’s future prospects, but when you’re a value investor you’ll often find yourself in uncertain situations where the uncertainty is offset (hopefully) by an attractively low share price.

This wasn’t a punt on an entirely unknown entity though. AstraZeneca had a long history of dividends and dividend growth, its balance sheet was strong and it was steadily increasing its R&D budget to fund the development of new blockbuster drugs.

So I invested, both in spite of and because of the obvious risks.

Holding: AstraZeneca struggles with the patent cliff

One of my favourite quotes about starting a company is that it’s like jumping off a cliff and trying to build an aeroplane on the way down.

That’s also a good description of AstraZeneca because by 2015 AstraZeneca had well and truly jumped off the patent cliff and was desperately trying to build its way out of the problem on the way down.

Rather than building an aeroplane, AstraZeneca’s survival plan revolved around its ability to research and develop new products to replace the old. Its plan involved large numbers of talented humans, powerful computers and advanced laboratories, and all of that costs a lot of money.

Despite the complexity of the underlying R&D activities, most of which are entirely beyond me, from a financial point of view it’s pretty simple:

Assuming at least a reasonably stable level of R&D productivity, the amount of money spent on R&D and the expected profits from that expense are quite closely related. In other words, if AstraZeneca doubles its R&D spend, it can expect (although not guarantee) about double the future profits.

The problem is that AstraZeneca has to spend money today in order to generate revenues and profits ten or twenty years from now. This is a problem that all capital-intensive companies face (although R&D is typically recorded as an expense on the income statement rather than a capital investment on the balance sheet, it is as an investment in future patents which can be seen as a kind of capital assets – i.e. an asset which will be productive for more than a year).

If money grew on trees then this wouldn’t be a problem, but the money for R&D has to come from somewhere and typically it comes from cash generated by today’s sales.

But what if those sales are falling because existing products are going off-patent? The cash still has to come from somewhere and the obvious alternative is to borrow it. And that’s exactly what AstraZeneca did.

Using debt to fund R&D and dividend payments

When I invested in 2015, Astra’s total borrowings were £7.0 billion. That’s about where its borrowings had been since 2007 when it took on about £7 billion of debt in order to acquire MedImmune in 2007.

With average profits at the time of £4 billion or so, that gave the company a relatively prudent debt-to-profit ratio of 1.7, a long way short of my maximum “prudent” value for defensive sector stocks of 5.0.

Since then, AstraZeneca’s borrowings have grown to more than £13 billion. That’s about six times its recent average profits of just over £2 billion, so I would no longer consider the company’s borrowings prudent.

On the plus side, those additional borrowings helped fund a massive increase in R&D spend at a time when profits and cash flows from operations were stalling. The borrowings also helped the company avoid either a rights issue or a dividend cut, which are the typical “last resort” ways to raise cash.

You can see this in the chart below, which shows a) increasing R&D, dividends and borrowings (up 26%, 22% and 90% respectively), and b) virtually flat revenues and profits (up 4% and 8% respectively).

Selling: The share price goes up as Mr Market becomes more optimistic about AstraZeneca’s future

Having spent more than £30 billion on R&D in the last decade, it seems as if AstraZeneca could be developing a pipeline of products capable of offsetting the loss of its previous blockbuster drugs. For example, two relatively new products, Brilinta and Farxiga, have both recently broken the $1 billion sales barrier required to achieve “blockbuster” status.

As investors have become more optimistic about AstraZeneca’s future (although admittedly this optimism is very fragile), they have been willing to pay a higher price for its shares. The result is an increase in the share price of more than 15% over the last couple of years.

That isn’t much in the grand scheme of things, but when combined with AstraZeneca’s weak financial results (remember, Mr Market is becoming more optimistic about the future, but AstraZeneca’s recent results are still unimpressive) the company’s growing valuation ratios leave it with a fairly poor rank on my stock screen.

So who’s right? The increasingly (but inconsistently) optimistic Mr Market, or my more pessimistic stock screen? Of course, nobody can really know, but personally, I’m at the less optimistic end of the scale.

Why? It’s partly because AstraZeneca’s dividend yield is currently about equal to the FTSE 100’s yield. The implication is that investors think AstraZeneca can increase its dividend over the long-term more quickly than the index (AstraZeneca is more risky than the FTSE 100, so if their yields are equal then investors must expect higher growth from AstraZeneca as compensation for the increased risk).

Is that likely? Again nobody can know, but I do know that the company has taken its debts from £7 billion to £13 billion in just a couple of years, largely (in my opinion) to fund additional research and development. It had to borrow the money because its operations weren’t generating enough cash to fund the R&D increases and the dividend. So what happens next?

What if AstraZeneca has to keep borrowing in order to fund ever greater R&D as it attempts to offset collapsing profit margins caused by the patent cliff?

I’m sure AstraZeneca could add a few more billion to its debt pile, but all that will do is make the company more and more risky.

If it all works out over the next few years then the CEO will be hailed as a hero, AstraZeneca’s share price will probably shoot up and I’ll be branded (perhaps correctly) as an idiot.

If it doesn’t work out, a dividend cut or rights issue may be needed to reduce borrowings. The company could also suffer a permanent reduction in size if its pipeline of potential drugs turns out to be full of nice ideas that fail to generate actual improvements in medical outcomes, and therefore fail to generate revenues and profits.

Either way, I think it’s likely that AstraZeneca’s higher share price and lower dividend yield no longer offer a sufficient margin of safety from these risks.

As a result, I have removed AstraZeneca from the model portfolio and my personal portfolio. The cash proceeds will, as usual, be reinvested into a new holding next month.