When I switched from deep value investing to defensive value investing in 2011, the first stock I purchased with the new approach was BP.
Six volatile years later, this has turned out to be a mildly disappointing investment. That’s partly because of bad luck, but it’s also because I paid too much in the first place. That’s a mistake I don’t plan on making again.
Here are the headline results:
- Purchase price: 494p on 04/03/2011
- Sale price: 529p on 05/01/2018
- Holding period: 6 years 10 months
- Capital gain: 7.1%
- Dividend income: 34.0%
- Annualised return: 5.9%
Back in 2011 my defensive value investing rules were barely formed and much has changed in the intervening six years, but the basics were there. I was looking for consistent dividends and profits, consistent growth, combined with high profitability, low debts and an attractive share price relative to ten-year average earnings.
BP certainly had most of those features in 2011, with revenues, profits and dividends growing steadily and rapidly since the beginning of the 21st century. Debts were reasonable, returns on capital employed were good and the share price was low as a result of the 2010 Deepwater Horizon disaster.
So I invested, and the result was a volatile, slightly disappointing, but not entirely fruitless investment:
Buying a cyclical company too near the top of its cycle (again)
As with Braemar Shipping and Rio Tinto, both of which have left the model portfolio in recent months, the late 2000s were a time when BP was riding high on the back of the commodity super-cycle. When BP joined the portfolio its track record reflected the huge run-up in oil prices since the turn of the millennium:
The chart also shows the near-total reduction of the dividend in 2010 which was one of the early financial impacts of the Deepwater Horizon disaster. The chart also ignores BP’s financial liability relating to the disaster, which at the time was estimated to be close to $40 billion.
Obviously, $40 billion is a huge amount, but I wasn’t overly worried. BP’s then-recent average post-tax profits came to almost $19 billion, so the estimated liability was “only” a little over two years’ profit.
My assumption at the time was that BP would be able to cover the financial costs without too much trouble. The usual array of options would be open to management, from increasing efficiency (i.e. reducing costs) to re-focusing on the core business (i.e. selling off non-core assets) as well as dividend cuts and rights issues.
So despite the disaster-related cloud hanging over BP at the time, I was happy to invest.
Holding on as BP sold off assets and faced an oil price crash
In the years following the Deepwater Horizon disaster, BP’s disaster-related financial liabilities increased from around $40 billion to a final figure of just over $60 billion. As expected, BP has so far funded much of this liability by selling non-core assets, raising more than $50 billion to date. As a result, BP has shrunk, but that was the least of its problems.
A much bigger problem came from something far less controversial and far more predictable than the Deepwater Horizon disaster: The price of oil.
It may seem ridiculously obvious to say that an oil exploration and production company like BP is highly sensitive to oil prices, but it’s a fact I woefully under-appreciated when I bought BP in 2011. Put simply, companies like BP have lots of relatively fixed costs (oil rigs, refineries, pipelines, etc.) and virtually no ability to set the price of the product they sell.
For example, if the price of oil goes from $20 to $150 (as it did between 1999 and 2008) then companies like BP are going to make huge profits even if the CEO is a complete idiot. If the price of oil goes from $150 to $30 (as it did between 2008 and 2016) then companies like BP are going to see their profits fall, even if the CEO is the greatest genius on the face of the earth.
Even worse, high oil prices and good results over several years are often followed by lower oil prices and bad results over several years. That’s because high oil prices and high profits typically lead to an increase in oil supplies as companies desperately try to find and extract more of that black gold (or partial alternatives such as natural gas).
It may take a while, but once supply increases then economics 101 says that prices will decline. And of course, once oil prices decline, profits decline. Boom turns to bust and investors who expected the glory days to last forever are left high and dry when share prices collapse.
This is exactly what happened to BP. When the oil price declined from $100 in mid-2014 to $30 in early 2016, BP’s shares went from about 520p to less than 340p. Remember, this is one of the most analysed stocks on the planet, so I think it’s safe to assume that most investors (including me) failed to see the oil price decline coming.
While a falling share price can be unpleasant, it shouldn’t really matter as long as the company can grow its revenues, earnings and dividends over the long term. If the oil price decline is reversed fairly quickly then its impact on profits may be negligible. But that isn’t the case here because the price of oil is still “only” $60, far below its $100 price tag from 2011.
As a result, BP’s earnings for the past two years have been much lower than in previous years. That’s true even if you look at “replacement cost profit”, which excludes the impact of falling oil prices on the value of the company’s existing oil inventories (reported profits, which include the impact on inventories, have been negative for the last two years).
My point here is that while BP has undoubtedly done lots of good work over the last few years, sucking untold amounts of oil out of the ground so that we can burn it for fuel or turn it into plastic, that was not the most important thing. The most important thing was the oil price, a factor which is both volatile and unpredictable.
You can see the overwhelming importance of the oil price in the following chart, which shows BP’s financial results up to 2016 (using the more optimistic “replacement cost profit” figures for 2015/2016):
Selling because growing debt and pension liabilities make BP a highly uncertain investment
If BP’s falling earnings were the only problem then I might have been willing to continue holding, waiting and hoping for an oil price recovery which would underpin the sustainability of the dividend.
To be honest, that’s the approach I’ve taken for much of the last few years, during which time BP has consistently been one of the least attractive holdings according to my stock screen.
But BP now has other problems too, which have become serious over the last year or so. Chief among them is its growing financial obligations in the form of debt and pension liabilities.
When I bought the company in 2011 it had borrowings of about £29 billion, which was high relative to the previous five years where borrowings had averaged £16 billion. Today, despite earnings which are far lower than they were back then, BP’s borrowings have ballooned to more than £47 billion.
This gives BP a debt ratio (the ratio of debt to five-year average post-tax profits) of 11.3 using reported profits or 7.5 using replacement cost profits. Either way, the ratio is far above 4.0, which is the most I’m comfortable with for cyclical stocks.
I don’t know for sure, but I have a sneaking suspicion that these debts have been piling up so that the dividend can be maintained despite weak profits. If that’s the case then it’s a dangerous game for the CEO to play as this is a highly unsustainable strategy.
Another problem is BP’s defined benefit pension obligation, which has grown from £27 billion in 2011 to £40 billion today. This liability, combined with the company’s borrowings, gives BP a debt plus pension ratio (the ratio of borrowings plus total pension obligations to five-year average post-tax profits) which is far above my preferred maximum of ten.
The pension fund also has an almost £7 billion deficit, which is effectively another form of debt. Adding that deficit onto the company’s existing borrowings gives BP a debt plus deficit ratio of 8.5 or 12.3 (depending on whether you use reported or replacement cost profits), which is higher than my preferred maximum of 4.0.
So not only is BP a highly cyclical company, it’s also one which is loaded up with financial obligations, and that is not a combination I want in the model portfolio.
Lessons learned from a highly cyclical investment
There are no new lessons from BP. Instead, the lessons are the same as those learned from the Braemar Shipping and Rio Tinto investments. The key lesson relates to what is an attractive purchase price.
For less cyclical companies such as Unilever (an extreme example), a higher price relative to past earnings and dividends may be acceptable because there’s a good chance those earnings and dividends will be higher in the next decade than they were in the last decade.
So for Unilever, a purchase price which is 20 or even 30 times the average earnings of the last decade may only be ten times the average earnings of the next decade. In other words, the price may seem expensive relative to past earnings, but it may actually be cheap relative to (higher) future earnings.
For highly cyclical companies that logic is much less likely to apply.
If a highly cyclical company doubled its earnings over the last decade, it may seem reasonable to expect the company to double its earnings again over the next decade. And yes, it might, but it might also earn the same amount as in the previous decade, or it might earn less.
The fate of the company will depend on its industry cycle more than anything else.
I have no interest in trying to predict industry cycles. Instead, I would rather use strict valuation rules to limit purchases of highly cyclical stocks to periods where they are very likely to be near the bottom of their cycle.
This should help to reduce downside risks and increase the chances of the share price increasing whenever the cycle eventually turns upwards again.
This isn’t an exact science, but having crunched the numbers on a variety of companies these last few months, my rule for investing in highly cyclical stocks now looks like this:
- Only buy a company from a highly cyclical sector (Mining; Oil & Gas Producers; Oil Equipment, Services & Distribution; Home Construction) if its PE10 and PD10 ratios are below 10 and 20 respectively.
These limits of 10 and 20 for PE10 and PD10 respectively are far lower than my standard limits of 30 and 60. While 30 and 60 may be reasonable limits for Unilever and its defensive peers, they are not suitable for highly cyclical stocks like BP.
If I’d followed this new rule back in 2011 then I probably would have purchased BP at 380p in September 2011, rather than 494p in March 2011. That lower and more recent purchase price would have produced an 11.3% annualised return rather than a 5.9% annual return. That’s a fairly significant improvement in returns.
And because the purchase price would have been lower, the maximum share price decline (i.e. the moment of maximum stress for the investor and the point at which they’re most likely to panic-sell) would have been much lower. In fact, when the shares fell to about 340p in early 2016, the “paper loss” would have been a mere 10% or so compared to the actual decline of 30%.
So in summary, buying at a lower price would have increased returns, reduced risk and reduced stress. What’s not to like about that?