My recent article on FFP received a lot of attention, but most comments were along the following lines: “sure there is a 50% discount, but what if the discounts stays at this level?”. I would say: that’s exactly what’s value investing is about! You buy cheap, and you have to have some faith that at some point in time the market will recognize how much its truly worth. This is what Benjamin Graham had to say about the subject in 1955 to a Senate commission (the full transcript is available online):
The Chairman: … One other question and I will desist. When you find a special situation and you decide, just for illustration, that you can buy for 10 and it is worth 30, and you take a position, and then you cannot realize it until a lot of other people decide it is worth 30, how is that process brought about – by advertising, or what happens?
Mr. Graham: That is one of the mysteries of our business, and it is a mystery to me as well as to everybody else. We know from experience that eventually the market catches up with value. It realizes it in one way or another.
I’m not saying that the discount will completely be eliminated in time, because there are valid reasons why a holding company or a closed-end fund should trade at a discount. These are overhead costs, tax inefficiencies and potential value destruction. In general overhead costs for holding companies are low, while closed-end funds incur higher costs because there is an ‘active investment manager’ that gets a fee based on AUM. Holding companies on the other hand are often less tax efficient. Dividends cannot always be passed through from the operating companies to the holding company in a tax free manner. In most jurisdictions this is not a major issue: It is usually possible to pass through dividends in a tax free manner if the holding company has a sufficiently big stake in the operating company.
The last important ingredient that should determine the discount is whether or not the manager is adding value, subtracting value, or is value neutral. I think the best assumption is usually the last one. Most CEF’s invest for example in large cap companies: they don’t underperform the market before trading costs in the long-term, but they also don’t deliver outperformance. If you already account for trading costs under the overhead the manager probably doesn’t add further negative performance. Holding companies are very often passive vehicles, and since they are passive there is neither value destruction nor creation. Rella is a nice example of a holding company that is actually creating value, simply by buying back undervalued shares.
Discount fluctuations
Both CEF discounts and holdco discounts usually vary over time. Sometimes they are small, and sometimes they are big. My thesis is that these fluctuations can be partly explained by the perceived attractiveness of the underlying assets, general investor sentiment and the fact that both holdco’s and CEF’s have a fixed supply of shares outstanding.
I wrote for example about mortgage closed-end funds two months ago. Rising interest rates resulted in declining prices for the underlying assets and at the same time the number of investors willing to invest in CEF with these type of assets also declined. As a result the discounts to NAV increased at the same time as NAV’s declined. If the market would be efficient this would be hard to explain since the lower underlying asset prices should already have reflected the bad news.
The reason that this doesn’t happen is that there is no (easy) arbitrage between share price and NAV. So when investors want to exit anyway only one thing can happen: the discount increases. Similar mechanics are at work behind the holdco discounts, except that there is even less of a possibility of arbitrage between share price and underlying value. When a closed-end fund starts trading at a huge discount activist investors can usually acquire a big enough stake to push for liquidation while holdco’s are often tightly controlled by their founders. They might want to take advantage of the discount by buying back shares at some point, but it’s also possible that you simply have to wait for sentiment to change.
Possible returns
I have no idea when investor sentiment is going to change for FFP, but I do think it will probably change at some point in time. Between 2007 and today the discount varied between 30 and 50%, and in the next five or ten years I would be surprised if it wouldn’t move back to the bottom of this range at some point in time. And you shouldn’t underestimate the amount of alpha you can create when this does happen.
If it would take 10 full years for the discount to shrink from 50% to 30% you still generate 2.9% alpha annualized (this is after accounting for 0.5% in overhead costs). If it takes ‘just’ 5 years you are looking at 6.4% annualized alpha. I don’t know about you, but that’s certainly attractive enough for me!
Disclosure
Long FFP.PA and RELLA.CO
Hi, sorry to bother you again on FFP.
1. The “alpha” you mentioned is, of course, not really an “alpha”. You are forgetting about the opportunity cost (or financing cost). I do not know what your “benchmark” is, but let’s take the cash as an example: I am not sure if you would outperform cash on this 10yr horizon. The upside would then be (very) small, with taking a (big) risk that the NAV could go down – since you haven’t analyzed these ones in detail (not sure, but it seems to be the case given your responses).
2. If you want to refer to Benjamin Graham, fine, but since I assume you also read “Security Analysis”, you should know to analyze the underlying shares as well. According to “The Intelligent Investor” mr. Market is currently offering the underlying shares at current prices. But, what is their real value (!). You are now assuming the underlying shares are fair prices and mr. Market is right on that. But mr. Market is not right about the discount. I am very interested in how you see this potential inconsistency ?
3. If you really want to play the discount, you should short the listed underlying shares. Then you have a far better investment case. I guess this is basically what you are targetting !
Regards,
Pepijn (a fellow dutchmen)
Always happy to receive comments 🙂
1. The alpha is compared to the underlying assets, which is probably a decent proxy for the CAC 40 index. It certainly could be that the assets themselves deliver negative alpha.
2. It’s a potential inconsistency, but I think that in general the market is reasonable efficient, especially when we talk about the type of companies that FFP owns. I have indeed not done an in-depth analysis on the companies, but I obviously checked some key stats and I’m not that worried that they are extremely overvalued. You pay a 15-20x P/E on most of them, but they are also high quality businesses with high returns on equity and a good track record of growth. Peugeot is obviously a different case: cheaper, but crappy business. It’s not the kind of stuff I would normally buy, but I’m comfortable owning a basket of these stocks at a discount through another vehicle. It’s not like I’m buying junior gold miners or web 2.0 stocks at 200x PE ratio’s here.
3. That wouldn’t be a bad idea in a perfect world, but stock borrowing fees, margin requirements etc make it not so great. So I basically try to own the best leg of the trade you describe.
Concerning what would be a proper discount for FFP:
– The annual management costs are 10 MEuro. The value of the costs with 2% growth and 10% discount rate is 128 MEuro, which is about 6% of the NAV.
– Taxes paid seem to fluctuate wildly from year to year, so it is very difficult to assign a discount based on extra taxes caused by the holding structure.
The taxes in France are generally very high (E.g: capital gains 34,5 – 40 %).
The arbitrage argument is probably not valid here. FFP was founded in 1926. It has held at least the Peugeot shares so long that nearly all of the current market value is capital gain. Selling the positions would result into huge taxes. If the capital gain tax of 40% applies here, then selling the positions would eat up nearly the complete discount.
I think the easiest way to think about the negative value of the overhead is to express it in a percentage of the expected long-term return on the underlying assets. If you expect a 5% long-term return 0.5% in overhead would warrant a 10% discount, if you expect a 10% long-term return you would expect a 5% discount. Ok… maybe this was just a complicated way of saying I agree with your 6% number 😛
About the taxes: they certainly shouldn’t do something that is stupid from a tax perspective. If there are large capital gains taxes (probably not on all their holdings though) the easiest way to eliminate a discount would probably to just pass through 100% of all dividends received. I don’t think they would incur any additional taxes in that scenario.
Usually holdco discounts are around 20%. I think that should also in the FFP case be plenty to cover overhead costs and taxes.
Very interesting to read the various comments about NAV discounts. Its kind of a strange phenomenon how investors treat recurring earnings and cash flow as superior to asset value based values.
If a company were trading at a discount to its earnings power why don’t investors say, “the discount between a company’s price and earnings power may never close,”? Even in the examples of companies that have never returned capital to shareholders in any form are still bid up based on increased earnings. — I guess what I’m getting at is there is a sort of leap of faith in both instances that the market will magically see things your way.
Good point 🙂
Interesting angle. But, if you would do a real asset based valuation of a holding company, then you must (!) also include taxes they would have to pay if they sell the asset (i.e. what is the real value after costs). In this way you are consistent with the approach with regard to a “normal” operating entity.
Also typically, the shareholder friendliness with companies where there is a (very) large shareholder is simply small. This is simply the case more often with holding companies.
Why do you need to assume that the assets will be sold if you do an asset based valuation? An asset based valuation does not necessarily imply that we are looking at liquidation value. About the shareholder friendliness: it’s a good point, but at the same time companies that have low insider ownership have their own ways of screwing over shareholders. I rather have insiders own too much of the company than too little.
I am afraid you do have to take into account the taxes. In this case you should include a tax liability in your valuation.
If you want to do an asset based valuation or sum of parts, then basically you also would not want to value inventories for the full 100 pct or goodwill for that matter ! So in a correct approach would have to include the taxes.
I’m not saying that you should ignore taxes (that’s part of the reason that there should be a discount!), but taking the book value of the tax liability doesn’t make sense if that liability will only be triggered if the underlying assets are sold today. A liability that only needs to be settled when you want to settle it at some undetermined point in the future has a lot less (negative) value than the same liability today.
Great posts on CEFs.
We have been investing in CEF for years but only when we could see a catalyst occurring within the next year. The downside of holding many of these CEFs long term is the outrageous management fees. Do you know of any backtesting that has been done on CEF performance?
I’m not familiar with specific results, but if you Google there is a lot you can find :). Some interesting hits are these two articles, and if you follow the references you’ll get a quickly expanding reading list.
One last comment from me … Basically, when calculating an asset based valuation for a holding company you are essentially marked to market everthing. This simply means that if you take current prices as the best estimate of the underlying value of a stake then you should also deduct a tax liability from it to be consistent. As a value investor you should make conservative estimes of value and therefore include the tax liability at marked to market value as well.
We’ll see if this is your last comment 😉
But yes, you are marking to market everything for an asset based valuation. But simply taking the book value of the tax liability is not marking it to market!
Lets take this hypothetical example. A company has 1 million in debt with a zero interest rate that needs to be repaid in ten years time. The book value of the liability is 1 million, but what would be the value of this debt if it would be marked to market? I bet no-one would be willing to buy it at par given the zero interest rate, so marked to market it would be valued significantly below par.
For a deferred tax liability there is obviously no market, but that doesn’t mean that book value is the best estimate of market value. A deferred tax liability for a capital gains tax is much like the debt in the example above, except you don’t even have to repay it after 10 years if you don’t want to.
Thanks for this series of posts on FFP, after reading which I am now long the stock.
You have a very high quality blog.
Thanks! Remember not to trust everything I write though, I could be totally wrong… just check my latest post
Wait, doesn’t “Alphavulture” mean we beta vultures should just follow you blindly?