Journeys of a Bumbling Investor

On top of Joel Greenblatt's definition of ROC in the previous post, Morningstar's Pat Dorsey in his book "The Five Rules for Successful Stock Investing" defined Return on Invested Capital (ROIC) as:

Net Operating Profit After Tax (NOPAT) / Invested Capital

where

Invested Capital = Total assets – Non-interest-bearing current liabilities (e.g. accounts payable and other current liabilities) – Excess cash (cash not needed for day-to-day business needs)

Both definitions subtracts "excess cash" from the invested capital. But how does one determine the excess cash? how much of the cash figure in the balance sheet is excess?

One pretty decent way to determine excess cash is given in CFO magazine – see here and original article here.

To summarize, CFO magazine divided companies (of a particular industry) into quartiles, ranked by their cash/sales %. The required cash-level benchmark is the lowest quartile of cash/sales %. Each company's ratio of cash/sales is then compared with the industry benchmark to determine excess cash.

The formula is as follows: excess cash = [cash/sales % - benchmark cash/sales %] x [sales]. (If a company's cash level was below the benchmark level, it wasn't considered to have excess cash.)

Rating: Good

Key Points:

Joel Greenblatt proposed a “magic formula” to rank stocks based on 2 factors: Return on Capital (ROC), and Earnings Yield. Each stock will be ranked based on each individual factor (hence each stock has 2 ranks), then we add the two ranks up, and perform a final sorting. Hence the best score a stock can achieve, is 2.

Return on Capital tells us how good a company is, and is defined as follows:

EBIT / (Net Working Capital + Net Fixed Assets)

where

1. EBIT is used to allow us to view and compare the operating earnings of different companies without the distortions arising from differences in tax rates and debt levels.
2. It is assumed that depreciation and amortization expense (non-cash charges against earnings) are roughly equal to maintainance capex, hence it assumes that EBITDA – Maintenance capex = EBIT.
3. The “capital” in this case is the tangible capital employed. The tangible capital employed tells us how much capital is actually needed to conduct the company’s business.
4. ROA uses total assets, and ROE uses equity. Both total assets and equity values contain intangible assets and do not accurate reflect the amount of capital needed to do the business.
5. Net Working Capital is used because a company has to fund its receivables and inventory (excess cash not needed to conduct the business is not included in the current assets). But it does not have to fork out the full current assets, because it gets what is effectively an “interest-free loan” in the form of account payables. So current liabilities needs to be subtracted from current assets (except for short-term interest-bearing debt, which is money that the company still needs to fork out since its not free).
6. Net Fixed Assets (e.g. real estate, plant, equipment) is net of accumulated depreciation and additional capex.
7. Intangibles, specifically goodwill, is not included because we want to use “tangible capital employed”. Goodwill does not need to be constantly replaced, hence in most cases, return on tangible capital alone (excluding goodwill) will be a more accurate reflection of a business’ return on capital going forward.

Earnings Yield tell us how cheap a company is, and is defined as follows;

EBIT / Enterprise Value

where

1. Enterprise value = market value of equity (including preferred equity) + net interest-bearing debt
2. Note that non-interest-bearing debt (such as accounts payable) is not included in the enterprise value calculation as those are essentially “free money”.

Joel then recommends to hold 20-30 top ranked stocks, and turn them over once a year. A historical study over 17 years showed very good performance of around 30% annually compounded returns.

Thoughts:

1. I am not sure why in the formula quoted in the book for Enterprise value, cash & cash equivalents are not being subtracted. Not too sure whether or not that is intentional.
2. I would think that NOPAT would be a better measure than EBIT, to take into account the differing tax rates. The tax rate, unlike the debt level, is something that a company has no control over. Hence even though two companies may have the same “fundamental earnings generation capability”, the effective tax rate might pick out the winner.
3. We might need to include intangibles such as patents and such on top of the tangible capital employed, as these intangibles arise from R&D spending and needs to be constantly “replaced”. The way Greenblatt specifically singled out Goodwill as the intangible that needs to be excluded, and also that “in most cases, return on tangible capital excluding goodwill will be a more accurate ……..”, also implies that there are some intangibles that would need to be included.
4. I’m not sure why “non-current, tangible assets” are not included in the calculation of the tangible capital employed. Are there any such assets apart from fixed assets (i.e. PP&E)? Wouldn’t it be easier and a better catch-all to use (total assets – intangibles – excess cash) – current liabilities (excl. interest-bearing debt) instead? or (total assets – goodwill – excess cash) – current liabilities (excl. interest-bearing debt) depending on the scenario.

[Quick Reference: Current assets include cash, accounts receivable, inventory, marketable securities, prepaid expenses. Current liabilities include accounts payable, short-term debt, accrued liabilities.]

Another good article on Gurufocus.
According to Marty Whitman, the “safe and cheap” investor looks for four things in an investment:

• High quality balance sheet;
• Competent and shareholder-oriented management;
• Understandable and honest disclosure documents;
• Priced at 50 to 60 cents on a dollar.

Whitman is similar in many respects to Graham. Whitman concentrates more on the value of the company’s assets rather than the earnings that could be generated. Key quotes from the article:

• “It’s a lot easier to look at the quantity and quality of the assets and resources that a company has than to forecast its earnings. Assets can appreciate in value, can be enhanced, sold, or converted into something more productive.”
• “To Whitman, high quality means low debt, acreage of raw land, assets under management, fully paid rent, and other assets that can be easily valued.”
• “Marty Whitman thinks like a LBO control buyer. He asks: “How can I finance the transaction?” It is a balance sheet question regarding what can be put up as collateral to secure lending from the bankers.”
• “No more than 50 or 60 cents on the dollar for what a business would be worth to a private takeover buyer. It is absolutely crazy to pay more than 60 cents on a dollar for non-controlling interests in businesses. The outsiders always face agency problems.”

Rules of thumb for calculating his buying prices for various types of businesses:

• Financial-services companies and depositories: Stated book value.
• Small banks: 80% of book value.
• Mortgage portfolio: Calculate yield to maturity and perform credit analysis.
• Financial-guaranty insurers: Adjusted book value – a publicly disclosed number that is book value plus the equity in the present value of certain future premiums.
• Insurance companies: Adjusted book value.
• Real estate companies: Private appraisal value or market value.
• Real Estate (REITs): Appraisal value or discounted present value of cash flow from operations.
• Broker/dealer and asset managers: Tangible book value plus 2% of AUM.
• Operating companies: 10 times peak earnings or below “net asset value.”
• Tech companies: 2 times book value, less than 10 times peak earnings, 2 times revenue and cash larger than the book value of all liabilities.

I’ve also recently came across this video on The Ben Graham Centre for Value Investing (link here). There is also a nice interview here. Some more good takeaways follows:

Problems with his “safe and cheap” approach:

• In order to get a cheap enough price, typically the near-term outlook of the business has to suck. End up buying stocks that are 30 – 40 times earnings (i.e. earnings are really low, hence P/Es are high).
• In order to get businesses with huge balance sheets, you typically end up with ultra-conservative management. These companies don’t need access to capital markets due to their big strong balance sheets, hence they are non-promotional to Wall Street.
• Positions tend to suffer on marketability and liquidity.

Sell Discipline

• When we make a mistake, e.g. something happens that results in a threat of permanent impairment, competitive forces make the business not feasible.
• When securities become grossly overpriced.
• To meet redemptions (e.g. in 1998, 1998, value funds faced alot of redemptions)

Difference between Graham’s Net-Net and Third Avenue Funds’ Net-Net (from Third Avenue Funds’ 06Q1 Shareholder Letter):- Graham defines “net current assets”, “net working capital value”, “net net asset value” as (Working Capital – All Liabilities – Preferred Stock), which is equivalent to (Current Assets – Current Liabilities – All Liabilities – Preferred Stock). The value is a conservative measure of liquidation value, which makes the conservative assumption that at least enough can be realized from the plant account and miscellaneous assets to offset any shrinkage sustained in the process of turning current assets into cash. Graham noted that the purchase of a diversified group of companies on this bargain basis (i.e. below net-net value) is almost certain to result profitably within a reasonable period of time.

Martin Whitman uses the Net-Net concept with the following caveats:

• Whitman will only look at the Net-Net value if the company is extremely well financed, i.e. it can definitely meet its obligations to its creditors [Note: going-concern]. This is because the current asset value might be boosted up alot by stuff such as inventory, costs in excess of billings, receivables from less than credit-worthy customers, which would not help the company meet its obligations.
• Appropriate re-classification needs to be made on a case-by-case basis. Some current assets such as department store merchandise inventories, are actually lousy fixed assets. They will fetch much less than book value when sold in a Going Out of Business sale, and even as a going-concern, factors such as markdowns, obsolescence, seasonality, mislocation will significantly impact what you can sell them for. On the other hand, some marketable securities are not considered current assets in GAAP when they should be. If you have a fully-leased, Class A office building with credit-worthy tenants on long-term leases, that should be classified as a current asset (in fact, near cash), since you can sell the building very easily.
• Off-balance-sheet liabilities (e.g. may be disclosed in footnotes) need to be subtracted in the net-net calculation.
• Excessive expenses and losses need to be capitalized and included in the liabilities.
• Certain fixed assets, e.g. PP&E can sometimes create cash. E.g. sales of PP&E at a loss can generate tax benefits / cash refunds.

Just read an article on Gurufocus: http://www.gurufocus.com/news.php?id=877 that talks about Benjamin Graham's "net-net" investment strategy in 1979, and Joel Greenblatt's analysis of it back in 1981.

Salient points:
Graham's rough liquidation value (net-net) estimate:
"Current Assets" (cash, accounts receivable, inventory, etc.)
Less: "Current Liabilities" (short term debt, accounts payable, etc.)
Less: "Long Term Liabilities" (long term debt, capitalized leases, etc.)
Less: "Preferred Stock" (claim on corporate assets before common stock)
Divided by: Total Shares Outstanding
EQUALS "Liquidating Value Per Share"

Greenblatt did a study using 6 years of historical data (1972-1978) with the following 2 conditions:

1. Stocks that have shown a loss over the last 12 months, were not considered
2. Stocks were sold after 100% or 2 years (whichever came first)

The screen that worked best was

1. Price / Liquidating Value <= 0.85
2. Price / Earnings <= 5

which returned an annually compounded rate of 42.2% (before transaction costs).

Conclusion: You don't find such bargains (companies selling for less than net working capital) anymore =)