We’ve recently received several questions about our valuation methodology. Specifically, readers have asked why we include property, plant and equipment in our valuation, and why we only discount it by half, as opposed to a higher figure (two-thirds, four-fifths, one-hundred percent). They are concerned that by including property, plant and equipment in our assessment, or by failing to apply a sufficient discount to those assets, we are overstating the asset or liquidation value of the companies we cover and therefore overpaying for their stock. In this post, we better describe our approach to asset valuation. In the next post, we deal with our method for protecting ourselves from overpaying for stock.
Our valuation methodology is closely based on Benjamin Graham’s approach, which he set out in Security Analysis and The Interpretation of Financial Statements. Like Graham, we have a strong preference for current assets, and, in particular, cash. As we mention on the About Greenbackd page, our favorite stocks are those backed by greenbacks, hence our name: Greenbackd. We love to find what Graham described as gold-dollars-with-strings-attached that can be purchased for 50 cents. We believe that there is value in long-term and fixed assets, although not necessarily the value at which those assets are carried in the financial statements. The appropriate discount for long-term and fixed assets is something with which we (and we suspect other Grahamite / asset / liquidation investors) struggle. We think it’s useful to consider Graham’s approach, which we’ve set out below:
Graham’s approach to valuing long-term and fixed assets
Graham’s preference was clearly for current assets, as this quote from Chapter XXIV of The Interpretation of Financial Statements: The Classic 1937 Edition demonstrates:
It is particularly interesting when the current assets make up a relatively large part of the total assets, and the liabilities ahead of the common are relatively small. This is true because the current assets usually suffer a much smaller loss in liquidation than do the fixed assets. In some cases of liquidation it happens that the fixed assets realize only about enough to make up the shrinkage in the current assets.
Hence the “net current asset value” of an industrial security is likely to constitute a rough measure of its liquidating value. It is found by taking the net current assets (or “working capital”) alone and deducting therefrom the full claims of all senior securities. When a stock is selling at much less than its net current asset value, this fact is always of interest, although it is by no means conclusive proof that the issue is undervalued.
Despite Graham’s cautionary tone above, he did not necessarily exclude long-term and fixed assets from his assessment of value. He did, however, heavily discount those assets (from Chapter XLIII of Security Analysis: The Classic 1934 Edition “Significance of the Current Asset Value”):
The value to be ascribed to the assets however, will vary according to their character. The following schedule indicates fairly well the relative dependability of various types of assets in liquidation.
Graham then set out an example valuation for White Motor Company:
In studying this computation it must be borne in mind that our object is not to determine the exact liquidating value of White Motor, but merely to form a rough idea of this liquidating value in order ascertain whether or not the shares are selling for less than the stockholders could actually take of the business. The latter question is answered very definitively in the affirmative. With a full allowance for possible error, there was no doubt at all that White Motor would liquidate for a great deal more than $8 per share or $5,200,000 for the company. The striking fact that the cash assets alone considerably exceed this figure, after deducting all liabilities, completely clinched the argument on this score.
Current-asset Value a Rough Measure of Liquidating Value. – The estimate values in liquidation as given for White Motor are somewhat lower in respect of inventories and somewhat higher as regards the fixed and miscellaneous assets than one might be inclined to adopt in other examples. We are allowing for the fact that motor-truck inventories are likely to be less salable than the average. On the other hand some of the assets listed as noncurrent, in particular the investment in White Motor Securities Corporation, would be likely to yield a larger proportion of their book values than the ordinary property account. It will be seen that White Motor’s estimated liquidating value (about $31 per share) is not far from the current-asset value ($34 per share). In the typical case it may be said that the noncurrent assets are likely to realize enough to make up most of the shrinkage suffered in the liquidation of the quick assets. Hence our first thesis, viz., that the current-asset value affords a rough measure of the liquidating value.
Greenbackd’s approach to valuing long-term and fixed assets
The first thing to note is that we’ve got no particular insight into any of the companies that we write about or the actual value of the companies’ assets. The valuations are based on the same generalized, unsophisticated, purely mathematical application of Graham’s formula. Further, if the actual value of an asset is objectively known or determinable, then we don’t know it and, in most cases, can’t determine it. That puts us at a disadvantage to those who do know the assets’ real value or can make that determination. Secondly, we can’t make the fine judgements about value that Graham has made in the White Motor example above. Perhaps it’s blindingly obvious that “motor-truck inventories are likely to be less salable than the average,” but we don’t know anything about motor-truck inventories or the average. It’s specific knowledge that we don’t have, which means that we are forced to mechanically apply the same discount to all assets of the same type.
Given that we’ve disclaimed any ability to actually value an asset or class of assets, why not adopt the lower to middle end of Graham’s valuation range for those assets? (Editors note: What a good suggestion. From here on in, we’re taking Graham’s advice. It’s simply because, in our experience, as idiosyncratic as it has been, an 80% discount to property, plant and equipment is too much in most instances. We think that 50% is a conservative estimate. In our limited experience, commercial and industrial real estate rarely seems to sell at much less than 15% below book value, and that’s in the recent collapse.) At first blush, specialist plant and equipment might appear to be worthless because the resale market is too small, but it can also be sold at a premium to its carrying value. For example, in the recent resources boom, we heard from an acquaintance in the mining industry that mining truck tires were so scarce as to sell in many instances at a higher price second hand than new. Apparently entire junked mining trucks were purchased in one country and shipped to another simply for the tires. Without that specialist knowledge of the mining industry, one might have ascribed a minimal value to an irreparable mining truck or a pile of used mining truck tires and missed the opportunity. What these examples demonstrate, in our opinion, is that the sale price for an asset to be sold out of liquidation is extremely difficult to judge until the actual sale, by which time it’s way too late to make an investment decision.
The best that we can do is fix a point at which we feel that we a more likely to be right than wrong about the value but will also have enough opportunities to invest to make the exercise worthwhile. For us, that point is roughly 20% 50% for property, plant and equipment. That 20% 50% is not based on anything more than (Edit: Graham’s formula, which has stood the test of time and should be applied in most cases unless one has a very good reason not to do so our limited experience, which is insufficient to be statistically significant for any industry or sector, geographical location or time in the investment cycle.) We always set out for our readers our estimate so that you can amend our valuation if you think it’s not conservative enough or just plain wrong (if you do make that amendment, we’d love to hear about it, so that we can adjust our valuation in light of a better reasoned valuation).
We hope that this sheds some light on our process. We’d love to hear your thoughts on the problems with our reasoning.