I remain enthusiastic about the Third Avenue Value Fund’s (TAVF) current portfolio, the vast bulk of which seems best described as Graham & Dodd “Net-Nets” on steroids. However, it may be hard for investors to understand the TAVF investment approach unless the investor has some sense of the uses and limitations of financial accounting, especially as it pertains to value investing. Thus, in this quarterly letter I discuss “The Uses and Limitations of Financial Accounting for the Value Investor”.
The first rule to remember is that financial accounting can’t tell “The Truth.” Financial accounting systems such as Generally Accepted Accounting Principles (GAAP) have to be based on a relatively rigid set of assumptions that cannot in all (or even most) contexts reflect economic reality. What economic reality is must be determined by the user of GAAP (the analyst) not the preparer of GAAP financial statements (the CPA). Yet GAAP or other accounting systems such as International Financial Reporting Systems (IFRS) or Statutory Accounting Systems for U.S. insurance companies, are essential, irreplaceable, tools for the value analyst. GAAP and GAAP-like financial statements are virtually the only quantitative objective benchmarks available to the analyst in the vast majority of situations. The value analyst uses these objective benchmarks to ascertain his or her determinations of what economic reality is.
The second rule to remember is that financial statements, unadjusted by the user, almost always will be misleading in one context or another.
Accrual accounting, the basis for most of GAAP, tends to mislead because of the failure to reflect good estimations of the cash flow experience, whether positive or negative. (Financial textbooks such as Brealey and Myers’ Principles of Corporate Finance overestimate the importance of internally generated cash flows.)
Cash accounting tends to mislead because it doesn’t reflect any estimation of the wealth creation, or wealth destruction, experience.
The third thing to remember in value investing is that there is no Primacy of the Income Account despite its general acceptance on Wall Street, even by Graham & Dodd in their various editions of Security Analysis; Principles and Technique. Rather, every accounting entry can be important in a value analysis. There is recognition of the accounting cycle fact that every accounting number is derived from, modified by, and a function of, other accounting numbers. This grows out of the recognition by the value analyst that corporate value is created in a multiplicity of ways, sometime related to each other, sometimes not. At TAVF, for purposes of analysis, we recognize four different ways in which corporate values are created:
- Cash flows from operations available to securities holders. This, while not uncommon, seems rarer than most commentators seem to believe.
- Earnings, with earnings defined as creating wealth while consuming cash. This seems to be what the vast majority of businesses and the vast majority of governments do the vast majority of the time. For the vast majority of entities, it is hard for earnings to deliver good results to shareholders over the long term, unless the entity has access to capital markets or has created saleable assets. As a practical matter, for earnings to create corporate value, the entity has to remain, or become, creditworthy.
- Resource conversion results in the creation of corporate wealth through asset redeployments, liability restructurings, and changes of control. These are accomplished through mergers and acquisitions, liquidations, investments in new areas, cash purchases of control blocks of securities, voluntary exchanges, spin-offs, debt restructurings, repurchases of outstanding common stocks (either leaving the issuer a public entity or going private).
- Having access to capital markets on a super-attractive basis. Probably more corporate value has been created this way than any other, especially when such super-attractive access is combined with an ability for the entity to benefit from an absence, or relative absence, of income tax burdens. Such super-attractive access to capital markets includes both equity markets and credit markets. A primary example of an equity market where corporations had super-attractive access (free money if you wish) was the pre-2000 dot com bubble. Credit markets which are super-attractive include the availability of non-recourse mortgage loans for a long term at low, fixed-interest rates issued to finance the ownership of income-producing real estate.
A good example of the difficulty in prescribing GAAP rules that reflect economic reality revolves around accounting for financial instruments, whether those instruments should be carried at amortized cost less impairments, or at lower of cost or fair value (with fair value usually equaling market price). The Financial Accounting Standards Board (FASB) is considering the matter currently. FASB’s current position seems wise – disclose both amortized cost and fair value. One disclosure would be in the financial statements themselves and the other would be in the footnotes to the financial statements. The real-world scenarios for accounting for financial instruments are so complex, I’d leave it up to corporate managements to decide whether fair value ought to be in the financial statements or footnotes; ditto for amortized cost, of course. Whichever management chooses becomes a good tool to help value analysts appraise management, e.g., how promotional are they?
Take a look at the complexity involved in accounting for financial instruments. In appraising a portfolio of non-control, marketable common stocks, there is no question that the appropriate standard is fair value. In appraising a portfolio of high quality performing loans which the holder intends to hold to maturity, the preferred standard should be amortized cost. But even this is complicated. Assume, in the case of an equity portfolio, that the holder has designs on getting control of some of the issuers whose common stock is in the portfolio. Fair value here can be misleading since from the point of view of the holder, the lower the fair value (i.e., market price) the more favorable to the holder. Assume that the portfolio of performing loans is financed almost wholly by interest bearing liabilities, the collateral for which is the performing loan portfolio. Here the liabilities would be worth their principal amount, unless that holder had other resources which he could use to acquire the liabilities at a discount, if available. In this case, fair value seems a much better measure for the performing loans than amortized cost, simply because it is possible that an event of default on the performing loans will be measured by fair value rather than amortized cost.
It ought to be noted that concerning events of default, there exist two, not one, general tests of insolvency. First there is the balance sheet test – does the fair value of the assets exceed the claim value of the liabilities? In the above leveraged balance sheet, the fair value of the performing loan portfolio could have flunked the balance sheet test. Second, there is the income test – will the borrowing entity be able to service its liability obligations as they become due? Normally, in gauging insolvency, the income test is far more important than the balance sheet test. Maybe the performing loan portfolio would pass the income test with flying colors while flunking the balance sheet test. If so, I’d postulate that amortized cost were the more important measure; if not, I’d opt for fair value, However, it should be noted that whether or not an event of default exists would be determined by the covenants in loan agreements and bond indentures.
Note that fair value measures can have perverse effects for an analyst. Take portfolios of performing loans that are financed by liabilities that are not interest bearing. This is the situation that exists for insurance companies and pension plans. Most insurance companies and most pension plans are continually reinvesting money received from maturing obligations into new obligations and also investing new moneys into new obligations, the vast bulk of which will be performing loans held to maturity. Assume interest rates rise sharply. Fair values will decrease dramatically but future net investment income will be much greater than would have been the case if interest rates had not risen. Which should be weighed more heavily – the current reductions in fair value or the prospective increases in investment income? That is something for the analyst, not the accountant, to decide.
I learned a great lesson about inflation accounting in the late 1970s and early 1980s. At that time, the accounting authorities together with the Securities and Exchange Commission (SEC) put in place as a financial statement supplement, “Inflation Accounting,” i.e., Current Value Accounting. Inflation accounting was designed to ameliorate the effects of overstated earnings caused by companies in a highly inflationary environment taking inadequate depreciation charges for expenditures they would have to make to replace aging property and equipment. At the time, I thought this brought some truth to accounting, never realizing other effects of rampant inflation. Because costs increased so rapidly, there were huge benefits to many companies with large amounts of “sunk costs.” New competitors couldn’t afford to enter the industry and existing companies found it hard to expand. Net, net, probably competitive relief because of rampant inflation was more important than materially inadequate depreciation charges. There was no way GAAP could handle both.
It is much more important for the U.S. economy to have its accounting systems geared toward informing creditors in a meaningful fashion than it is to have accounting systems directed toward meeting the perceived needs of Outside Passive Minority Investors (OPMIs). First, there is a lot more credit outstanding in the economy than there is net worth. Second, creditors use accounting to help determine the creditworthiness of a company by estimating whether that company will be able to generate cash internally, both long- and short-term, to pay its bills, and by estimating whether that company is likely to have relatively continual access to capital markets, especially credit markets. In contrast, OPMIs tend to place overemphasis on one accounting number – reported earnings – in order to predict what stock market prices in the immediate future might be. Bluntly, accounting systems do not seem as if they can really be very helpful as a tool for predicting near-term equity prices in OPMI markets. As far as I can tell, near-term market prices for common stocks in non-arbitrage situations will continue to be a “random walk”.
Cash payments are very different from stock options from a creditor’s point of view. Cash payments by a company can affect the creditworthiness of a business. Cash payments, therefore, are a company and a creditor problem. With minor exceptions, the issuance of stock options has no effect whatsoever on the creditworthiness of a company. Instead, stock options result presently, or prospectively, in the dilution of existing stockholders’ ownership interests. Stock options are not a company and creditor problem. They are a stockholder problem.
Those who think of options as an expense have it wrong, at least from the Company and Creditor points of view. Warren Buffett is quoted as saying, “If options are not a form of compensation, what are they? If compensation isn’t an expense, what is it? And if expenses shouldn’t go into the calculations of earnings, where in the world should they go?” Frankly, the Buffett statement is an overgeneralization, even though most finance academics seem to be wholly in concurrence. Stock options are not compensation from the points of view of the company itself, or its creditors. Stock options certainly are “compensation” when looked at strictly from the point of view of stockholders. The issuance of options results in present, or potential, dilution of common stockholders’ interests.
Assuming stock options are to be treated as a company expense, what should that expense be? Presently, the disclosure required under GAAP assumes that the cost of options to the company equals the theoretical value of the options to the recipients. However, it is utterly ludicrous to suppose that the value of a benefit to a recipient has any necessary relationship to the cost to a company to bestow that benefit. It is as if a sales clerk who has a 40% off employee discount buys a $100 sweater from her department store for $60 and the store then states that it incurred a cost of $100 because that is what the sweater is worth to the clerk, even though the company’s actual cost for the sweater might be, say, $35. The real cost of an executive option to the company (rather than to its stockholders) equals the present value of the probability that option program will reduce the company’s future access to capital markets, especially equity markets. I would not know how to measure such a cost. In fact, there should be an offset to this cost, namely the present value of the probability that the options program increases the retention of talent and/or motivates that talent’s productivity.
The cases where stock options become a company problem as well as a stockholder problem seem few and far between. Options are a company problem long-term, insofar as they either cause the company to pay out cash (or property), or, if their issuance reduces access to capital markets. In general, those cases where stock options become a company problem seem to encompass the following:
- The company is committed, or required, to pay out an ultra-high percentage of future earnings as cash dividends to common stockholders. Such companies include Real Estate Investment Trusts (REITs).
- The potential issuance of common stock through the exercise of stock options reduces the company’s future access to capital markets to raise new funds.
- The company is committed to having the amount of common stock outstanding relatively fixed, and therefore, acquires for cash, or property, enough outstanding common stock to cover the new issuance of common stock through the exercise of stock options.
Over 80% of the TAVF portfolio is invested in the common stocks of companies I like to think of as Graham and Dodd Net-Nets on steroids. The reason for steroids is that we believe the prospects seem bright that over the next five-to-seven years, the net asset values (NAVs) of the TAVF portfolio companies will increase by not less than 10% to 20% per annum compounded. TAVF Net-Nets count as current assets high quality assets, surely convertible to cash in a year or so. Graham and Dodd, in contrast, define current assets only as assets conventional balance sheets.
Other things being equal, the Graham and Dodd Net-Nets held by TAVF are susceptible to resource conversion activities, especially changes of control, and going privates, i.e., insider buyouts. This is so because the common stocks are so cheap relative to business values and each company enjoys a super-strong financial position. I have no question that TAVF’s Hong Kong-based Net-Nets would be highly attractive to various Chinese mainland entities. However, there will never be a change of control in any of our Hong Kong Net-Nets unless it results from a friendly, negotiated transaction. Insiders in each company own enough common stock to thwart easily any hostile takeover attempts. If there ever were to be a change of control in any of the Hong Kong Common Stocks owned by TAVF, the prices paid for these securities ought to reflect huge premiums over current prices.
While going-private transactions almost always are priced at substantial premiums over then existing market prices, this is far from always attractive for buy-and-hold investors, such as TAVF. This is because the buyer in the going-private transaction has complete control of the timing of the transaction. The buyer has incentive to propose a transaction when market prices for the common stock are depressed.
I have no idea as to whether or not changes of control or transfers of elements of control, or going privates ever will occur for any the Net-Nets held in the TAVF portfolio. Rather, a basic reason for owning these common stocks is that we believe the prospects appear to be so very good that NAVs will grow over the next five-to-seven years at rates of not less than 10% to 20% per annum compounded.
The Fund’s common stock portfolio is invested in the issues of extremely well-capitalized companies that were acquired at prices that, at the time of acquisition, represented meaningful discounts to readily ascertainable NAVs. The NAVs became readily ascertainable insofar as the specific assets consisted of certain marketable securities: income-producing real estate; land suitable for development; and intangibles, such as mutual fund assets under management. Rarely (except for cash and equivalents) were these readily ascertainable asset values classified as current assets under GAAP. The Fund’s definition of Net-Nets is taken from Graham and Dodd’s Security Analysis, but with a few twists. Graham and Dodd relied on a GAAP classified balance sheet to define current assets in order to ascertain if a common stock was a Net-Net. TAVF uses its own judgment, rather than GAAP classification, to define current assets in order to decide what is a liquid, i.e., current asset.
Graham and Dodd describe Net-Nets in the 1962 edition of Security Analysis on pages 561 and 562:
We feel on more solid ground in discussing these cases in which the market price or the computed value based on earnings and dividends is less than the net current assets applicable to the common stock. [The reader will recall that in this computation we deduct all obligations and preferred stock from the working capital to determine the balance for the common.] From long experience with this type of situation we can say that it is always interesting, and that the purchase of a diversified group of companies on this ‘bargain basis’ is almost certain to result profitably within a reasonable period of time. One reason for calling such purchases bargain issues is that usually net-current-asset values may be considered a conservative measure of liquidation value. Thus as a practical matter such companies could be disposed of for not less than their working capital, if that capital is conservatively stated. It is a general rule that at least enough can be realized for the plant account and miscellaneous assets to offset any shrinkage sustained in the process of turning current assets into cash. [This rule would nearly always apply to a negotiated sale of the business to some reasonably interested buyer.] The working capital value behind a common stock can be readily computed. Consequently, by using this figure (i.e., Net-Net asset value) as the equivalent of ‘minimum liquidating value’ we can discuss with some degree of confidence the actual relationship between the market price of a stock and the realizable value of the business.
While Graham and Dodd seem to have invented the idea of Net-Nets, TAVF uses that idea with a number of modifications. First, the Fund is not interested in Net-Nets unless the company is extremely well-financed. A large quantity of current assets, especially if they consist of inventories, costs in excess of billings, or receivables from less than creditworthy customers, probably cannot help the common stock of a company which cannot meet its obligations to its creditors. Second, many current assets classified as current assets under GAAP are really fixed assets of the worst sort. Take department store merchandise inventories. If the department store is to be liquidated, merchandise inventories are indeed a current asset, convertible to cash within 12 months at prices that conceivably could be close to book value, although much less than book value may be realized if the merchandise is disposed of in a Going Out of Business sale. On the other hand, if the department store is a going concern, merchandise inventories are a fixed asset of the worst sort. The merchandise inventories have to be replaced, are hard to value, and are subject to markdowns, obsolescence, shrinkage, seasonality and mislocation. The Toyota Industries portfolio of marketable securities and the Brookfield Asset Management portfolio of Class A Office Buildings seem to be much more of a current asset than department store merchandise inventories even though, for GAAP purposes, Toyota Industries’ marketable securities, and Brookfield Asset Management’s Class A office buildings, are not considered a current asset. Third, the Graham and Dodd formulation does not account for off-balance sheet liabilities, which may or may not be disclosed in footnotes; nor, do Graham and Dodd take into account excessive expenses or losses. At TAVF, such expenses or losses are capitalized and added to liabilities. Fourth, Graham and Dodd only seem to recognize partially that certain fixed assets, e.g., property, plant, and equipment, can sometimes create cash. For example, under Section 1231 of the U.S. Internal Revenue Code, the sale at a loss of such assets used in a trade or business, usually gives rise to an ordinary loss for income tax purposes. In that case, a corporation may be able to apply the loss first to reduce the current year taxes and any excess loss might be used to get “quickie” cash refunds from the IRS with regard to taxes paid in the prior two years (sometimes five years).
The identification of Net-Nets has not proved that difficult for the Fund, even though most of TAVF’s investments now are outside the United States.
For most market participants, the most important accounting number is earnings per share reported for a quarter. There are times in corporate analysis where quarterly earnings deserve great weight. Those times exist for not-well-capitalized companies that need relatively continuous access to capital markets. Also, quarterly earnings can have a major impact on heavily margined portfolios. Neither of these conditions pertains to TAVF, which invests only in the common stocks of very well-capitalized companies; also, TAVF does not borrow funds. Quarterly earnings are highly important to TAVF, insofar as they provide evidence that a business has suffered, or is suffering, a permanent impairment. None of the current common stocks in the TAVF portfolio seem to be close to suffering a permanent impairment. Current tough economic times –especially– in North America and Europe, notwithstanding, Fund Management keeps a wary eye on probabilities of permanent impairment.
There are investment areas where financial accounting is important, indeed, vital. For example, financial statements seem always to be a key in credit analysis. However, there are equity areas where financial statements become relatively unimportant. These areas include equity participations in new inventions and new discoveries; and, also, certain areas where issuers are sitting on huge amounts of unrealized appreciation, which unrealized appreciation is almost impossible to measure by any tool available to a non-control investor. Those seem to be areas where value investors ought to fear to tread. Unlike most general market participants, value investors are bottom-up participants, people who aren’t interested in all securities. Many of such securities are just unsuitable and/or unanalyzable using value investing techniques.