Month: September 2016

Follow up on my GMO post

My recent post (here) on energy companies has led to a lot of interesting responses from readers. One reader has forwarded research by JP Morgan showing that most oil companies they track cover their dividends at around a price of USD 50 per barrel and, therefore, my claim that the currently high dividend yield is unsustainable is wrong. (Interestingly, most commenters have focused on the dividend yield, ignoring my arguments about the other two valuation metrics employed by GMO – a sign of the times?).

Maybe a clarification is in order here to avoid misunderstandings. I never said that every oil company has to cut its dividend. My statement was about the (market-weighted) average that is used by GMO to construct the index referred to in the post. I have chosen Royal Dutch because I was of the opinion that given its 200 billion market cap and high dividend, it is to some extent representative and therefore a good example.

No doubt, there will be companies whose dividends are sustainable at the current price level. Obviously, this doesn’t mean it is true for the sector overall.

Lukoil, for instance, with which I am somewhat familiar due to my bond position (here), looks as if its high dividend yield is fine at current levels. But: Lukoil is exceptionally well positioned on the cost side. Not only is Lukoil a traditional low-cost producer (remember this chart), but it has enjoyed falling costs courtesy of a heavily depreciated Ruble exchange rate. I do not see how Royal Dutch, which sports a similar dividend yield as Lukoil, can match this cost advantage in the near future, even if they are likely to get concessions from suppliers. Either Lukoil’s dividend is supersafe, or Royal Dutch’s is not. I believe the latter. Remember: Lukoil’s dividend yield is high because of the famous Russia discount, not necessarily because of a general undervaluation of the sector.

Apart from that, the responses have triggered my interest and I have decided to have a quick look at the matter. As I do not have access to JP Morgen research, I have looked what Goldman has to say on this issue. In a recent research piece they had this chart,


The chart shows the current dividend yield and the oil price at which Goldman deems the dividend sustainable for the most important (in terms of market cap) players without resorting to debt financing and assuming that their capex estimate is sufficient to maintain production volumes (“organically”). Unsurprisingly, the lower the dividend yield, the lower the required oil price to maintain it. Also note that Royal Dutch needs an oil price of 63 USD – a mere 50 percent above current levels – to keep paying its high dividend over the long – term. The same is true for other major oil companies such as BP, Total, etc.

I would say this confirms my analysis.

The companies whose dividends Goldman thinks safe at current oil price levels, have yields of between 2.5 and 4.4 percent. That’s not bad, but if these yields excite you, I would urge you to have a look at “undiscovered” gems Microsoft (dividend yield: 2.75%) and Cisco (3.34%), both of which I own (here).

Ok, Royal Dutch is doomed, what about the others? Surely, they are good value at nice yields plus a “free option” should oil prices rise?

There is no denying that a 50 percent higher oil price would help – but that’s (bullish) forecasting. Let’s work with what we have and play around with Goldman’s analysis. Let’s focus on the term “organically”.

As explained before, organically in this context means that the capex estimate Goldman uses to estimate the sustainable dividend is high enough to keep current production volumes indefinitely by replenishing reserves 1:1. This is necessary if you are valuing the company as a mature dividend payer. In case Goldman’s capex estimate is too low, your option betting on a higher oil price is not “free”, but paid for by reducing the capital stock. Unfortunately, Goldman is a little unspecific on how they estimate their sustainable level, but they seem to be guided management.

So we have to work around the problem to see whether this is a good idea.

Exxon dominates the sector with its 350 billion market cap. Goldman think’s its yield at 3.5 percent sustainable at current prices. It is considered to be one of the better managed majors, free from the type of government meddling you get with European integrated oil companies. Warren Buffet was a major shareholder until recently calling it a “wonderful business”. In short, the best management.

Nevertheless, despite doubling their annual capex during the oil boom, their reserve replacement ratio has been below 100 percent, if one excludes their acquisition of shale gas play XTO in 2009. That’s according to Jim Chanos (here) who called the integrated oil companies value traps and liquidating trusts for this reason. I observe: even with record capex in the past they were not able to replace their reserves organically.

Now, at lower oil prices capex will have to fall. Who knows, maybe less cash at hand will help focus the mind of management, but it is a bet you have to be willing to make. Also new reserves will likely only come from politically unstable regions (such as Exxon’s Kashgan field in Kasachstan) a risk you also have to factor in. I feel that at current oil prices there is not much margin of safety to compensate for these risks. Cisco’s dividend is safer.

Further, over the weekend there has been an interesting development regarding Exxon mobile’s book value. From the FT (here),

(…) Last week, it emerged that Eric Schneiderman, New York attorney-general, was investigating Exxon’s decisions not to take large charges to its profits for writedowns in the values of its assets following the fall in oil prices. That move built on the probe he launched last year into the company’s statements on climate change.

The SEC is now also looking at Exxon’s reporting of its reserves, asset valuations and writedowns, as well as its disclosures on the risks that climate change creates for its business (…) 

Wow, the SEC is questioning Exxon’s asset valuation, i.e. its book value – although the stock price is trading near its all-time high and nobody has lost any money.

And for a good reason: Management at Exxon, you know the guys who have spent billions on capex when oil was at 100 without actually finding much, think that a halving of the oil price has no consequences for their book value. And what’s better: they seem to be the only ones.

I mean, even Royal Dutch has impaired their assets by USD 8 billion, or 4(!) percent of their asset value. The same goes for Chevron, which at least has reduced book values by USD 4.7 billion.

Message: do not trust these managements (and book values).


Clearly, the oil industry still believes that oil prices will recover significantly. As do many investors who either think the dividend yield is sustainable, or the optionality in case of a higher oil price is worth a lot. This view is embedded in the valuations and therefore this is not a contrarian play. The sector (average stock) is not cheap. Moreover, book value, one of value investor’s favourite metrics appears to be unreliable, just as it was in 2007/2008 with the banks. If oil prices do not recover substantially, we are going to find it out.


They really don’t know what they are doing: QE and Spanish banks

I wrote on how negative rates are destroying the Italian banking system (here). Now, there is a new piece by Exane (here, via valueandopportunity) on how negative rates might play out in Spain. Unsurprisingly, the effects are similar.

Just as in Italy, Spanish banks have heavily invested in government bonds in order to benefit from the free central bank put provided by QE and whatever it takes. Just as in Italy, private sector lending has been crowded out (it probably was too high to begin with) and, just as in Italy, the low hanging fruit has been picked, now that Spanish government bonds yield less than the 10 year US treasury.

The report is entertaining to read, but the most interesting information can be found in this table,


It shows what percentage of the respective bank’s profit before tax (PBT) comes from interest payments and capital gains in the sovereign portfolio. The percentage is lowest for Santander (SAN) and BBVA, as they have substantial operations abroad and had not engaged in foolish real estate lending to begin with. Nevertheless, their dependence on income from their sovereign bond portfolio is substantial. For others, the importance of the sovereign book is way higher. For the seven largest Spanish banks, on average, 93 (!!!) percent of pre-tax net income comes from gains and income on their sovereign books.

The report mentions that the average remaining duration of the sovereign book is between 3-4 years, which means that over the next half decade virtually all of the Pre-tax profit in the Spanish banking system will disappear, if interest rates remain where they are.

Macro economists, as employed by central banks, no doubt would answer that this is exactly the purpose: force banks to lend more and help the economy grow out of the high debt (circular reasoning being a requirement for a PhD in this field).

They forget that in order to lend more to the private sector a bank a.) needs equity and b.) needs to find enough suitable costumers.

As for the equity, most of the equity in Spanish banks consists of tax loss carryforwards, i.e. it is not really there and not available to cover losses – and everybody knows that. Everybody who is not a macro economist, that is.

As for the costumers, Spanish banks are losing their best and biggest clients to – you guessed it – the ECB and to the bond market. Ask yourself: why should big Spanish corporates (Initex, Endesa etc.) pay the overhead costs for such unnecessary things as underwriting departments and bank branches, if they can pay near zero in ETF and central bank dominated bond markets?

Answer: they don’t.

Thus Spanish banks are trying to find costumers by tricking Spaniards into credit card debt with opaque zero cost offers. If the only way you hope to grow your loan book profitably, is to screw your costumers, I would say that this is neither the way to grow your economy, nor sustainable.

There is no conspiracy – central banks really do not know what they are doing.

Why commodity companies might not be as cheap as they seem (and GMO is likely wrong)

Excellent piece by GMO on investing in commodity stocks (I do not know why they call them resource equities, probably for marketing reasons).

There are very useful long-term charts comparing investing in equities vs. investing in the commodity directly. They show that equities in commodity producers have delivered a significant positive return even in the absence of a secular commodity bull market. Further, they demonstrate that for financial investors it is really the only alternative, as the long-term returns from investing in futures are horrible, as a result of negative roll returns. The charts are very useful if you want to market a commodity equity fund.

However, the following statement irritated me,


According to the metric they use, a combination of normalized historical earnings, P/B and Dividend Yield, the valuation of the energy/metals sector has the lowest historical relative valuation on record only matched with 1998 the days of 10 dollar oil.

The reason I was irritated, is that recently when I looked at a few oil mayors, I felt that at the current price level their valuation was excessive and their dividend not sustainable. Moreover this was true across the board, whether it is Exxon mobile or OMV (Austrian integrated oil company).

Just look at Royal Dutch Shell, with a dividend yield of over 6 percent, a normalized (5y) PE of 11 and trading at a price book of 1.3. That’s looks cheap according to GMO’s chosen metric and, no doubt, represents a significant discount the S&P 500.

Just how reliable are these metrics in this case?

The dividend yield at over 6 percent, no doubt is sporty, but in 2014 the dividend cover was a meagre 1.3 whereas in 2015 it dropped to an absolutely unsustainable 0.2. In other words: in the recent past Royal Dutch has paid dividends by tapping the generous debt markets. Clearly, there could be one-off effects on net income that have distorted the dividend cover and we should take that into account.

This brings us to normalized earnings. Obviously, if you believe that the normalized PE of 11 is representative, you will not be overly worried by the low dividend cover of the last two years, as it implies a sustainable 9 percent earnings yield.

But how representative is it?

Let’s look at the main P&L categories to get a better feel for the figures.

Year Net Income EBIT EBITDA
2011 30.9 42.7 55.9
2012 26.7 37.7 52.3
2013 16.4 26.9 48.3
2014 14.9 19.9 44.4
2015 1.9 -3.3 23.5
Average (bn USD) 18.2 24.8 44.9

Unsurprisingly, the figures show that the degree of profitability of any integrated energy company heavily depends on the price of energy. Furthermore it can be seen that 2014, when oil prices started to drop from 94 dollar per barrel to 50 dollar at YE, was still a very good year courtesy of a high average oil price for the whole of 2014. So, for four out of five years used to normalize the earnings saw an oil price trading in the 80-100 USD range – pretty much 100 percent above where we are now.

Normalizing earnings in this case only gives you a good picture if you forecast significantly higher oil prices going forward. But this would involve bold forecasting of oil prices and even Jeremy Grantham contends in his piece that this is impossible.

Unless oil prices significantly increase from here, expect the dividend to be cut or debt to increase further.

Now let’s look at P/B the last metric GMO uses. No doubt, Royal Dutch at a P/B of 1.3 looks cheap relative to the market, but this should not be surprising in a market dominated by tech and service companies with their high (and mostly irrelevant) P/B ratios.

The biggest issue I have with this metric, however, stems from the fact that book value is not what it used to be, i.e. a pure and objective measure of how much has been invested in the business so far. Royal Dutch values most of its assets on an expected cash flow and DCF basis, i.e. it is forward-looking and subjective in nature. Currently they are implying a 6 percent discount rate reflecting the record low-interest rate environment. Put differently, the book value is an implicit bet on low rates.

For this reason I prefer another valuation metric. Given that I cannot forecast oil prices the most reasonable assumption is to take the average price for last year, the first full year of 50-60 USD oil and see what they can earn on that basis. As it seems that net income and EBIT figures for 2015 are somewhat distorted by impairments on assets, I will base my analysis on the EBITDA for 2015 which amounted to 23.5 billion dollars. I will be generous and assume that management is able to cut costs (after all not all profitability depend on the price alone) to deal with the new environment as well as squeeze some synergies from their recent acquisition and postulate that the normalized EBITDA at the current price level can be around 30 billion dollars, i.e. an improvement of 30 percent.

At an enterprise value of 350 billion this translates into a very high normalized EV/EBITDA ratio of slightly below 12. Remember, this is for a business which depends on higher commodity prices and substantial capex for sales growth. Furthermore, the nature of the business is such that constant investment is required to replenish reserves. Over the past 10 years, for instance, Royal Dutch has constantly invested between 150-200 percent of depreciation without significantly increasing its oil reserves (although they have increased their liquefied nat gas capacity), a period over which it has constantly been free cash flow negative.

In line with that, management expects capex spending of about 25 billion per annum for the for the forseeable future. All this results in an expected free cash flow to the firm of 5 billion USD, not even 2 percent of the value of the business. The valuation rests on energy prices rebounding substantially.


All three metrics used by GMO rely on a significant rebound in oil prices in order to be valid. I believe the case of Royal Dutch Shell to be representative to a large extent, at least for the large players. Mr. Grantham implicitly is forecasting higher energy/metals prices. Whether this will happen or not is anybody’s guess, but given that nothing has happened in China yet, I would not bet on it.