Mechanical Investing, Part I

From time to time on this web log, I have hinted at the possibility of using mechanical screens to build a profitable portfolio. Of course, the natural response is that one should have a deep understanding of both the company and the sector of the market in which that company exists to make appropriate and proper investment decisions. However, what if all the advice given by the investment advisors and financial planners is completely bogus. What if it is possible to screen the universe of public companies for certain attributes and simply buy the best of the companies passing that screen. What if it was possible to be successful doing so without even knowing what these companies do or sell. Let us take a pilgrimage for a while with the goal of finding a better way to invest.

Several mechanical techniques for identifying good companies depend upon the calculation of the Return on Invested Capital (hereinafter denoted as “ROIC”). The normal formulation of ROIC is as follows:

ROIC = EBIT* / TIC,

where

EBIT* is a modified Earnings Before Interest and Taxes, and

TIC is the Total tangible Invested Capital.

Of particular interest at this juncture is the total tangible invested capital. This quantity is occasionally defined to be:

TIC = NWC + NFA,

where

NWC is the Net Working Capital, and

NFA is the Net Fixed Assets.

In The Little Book that Beats the Market, Joel Greenblatt used a floor function to obtain the Net Working Capital based upon the total current assets, the excess cash, and the amount of non-interest bearing payables. The problem is that Mr. Greenblatt was relatively opaque about how to obtain a value for the non-interest bearing payables and was an obscurantist with respect to the excess capital. To explore the derivation of the value of the net working capital, consider that Mr. Greenblatt used the following formulation:

NWC = MAX[0, TCA – EC – NIBP],

where

TCA is the Total Current Assets,

EC is the Excess Cash, and

NIBP is the Non-Interest Bearing Payables.

If your source for market information provides a value for the non-interest bearing payables for the 10,000 or so available companies in the market then you are in reasonably good shape: that is, if they provide the correct number. Mr. Greenblatt alludes to the possibility that the non-interest bearing payables could be obtained as the difference between the total current liabilities and the interest bearing payables. This may be written as:

NIBP = TCL – IBP,

where

TCL is the total current liabilities, and

IBP is the interest bearing payables.

It turns out that the interest bearing payables almost certainly show up on the EDGAR filings of the company and are designated as either the short-term debt or the current long-term debt due and owing. We define the interest bearing payables as the sum of the short-term debt and the long-term debt currently due and owing. It is common accounting practice to define the total current assets and the total current liabilities as follows:

TCA = (C + STInv) + (AR + Inv + OCA), and

TCL = (AP + STD + OCL),

where

C is the total amount of cash held by the company,

STInv is the amount of Short-Term Investments held by the company,

AR is the accounts receivable,

Inv is the inventory belonging to the company,

OCA is the “Other Current Assets” belonging to the company.

Also,

AP is the accounts payable,

STD is the short-term debt or current long-term debt due and owing,

OCL is the “Other Current Liabilities” of the company.

Fortunately, all of these last eight quantities are available in either the 10Q or 10K filed with the Securities and Exchange Commission and these quantities may, almost certainly, be available in the commerical market databases.

Now, we return to obtaining the value for the non-interest bearing payables. Following the aforementioned formulation, we may write that:

NIBP = TCL – STD,

or,

NIBP = (AP + OCL).

The import of this last step in the derivation is not to be overlooked. It is one of the two principle means by which a mechanical screen, which uses the ROIC, can differentiate between a management team that is being prudent and efficient with management of the cash flow of the company and a management team that needs some improvement in this area.

We now turn our attention to obtaining a formulation for the excess cash of the company. It has bee suggested that the excess cash should be the difference between the cash and cash equivalents and the working cash of the company. That is, the excess cash may be defined to be:

EC = (C + STInv) – WC,

where

WC is the working cash of the company.

The working cash of the company is not to be confused with the net working capital of the company. The working cash is defined to be the difference between the liabilities of the company, which relate to the general operation of the company, and the assets of the company. In other contexts, the working cash of the company may be used to describe how far the company is inside of the door of the bankruptcy court. This quantity has been defined as follows:

WC = MAX[0, (AP + OCL) – (AR + Inv + OCA)].

Of greater import, this quantity is positive only when the current liabilities, without the short-term debt, exceed the current assets. It is possible that companies may exist with a non-zero amount of working cash but still be quite viable companies as investment opportunities. At this time, we are in a position of expanding the formulation for the net working capital given above to obtain a working version for this quantity. Specifically, we may write that:

NWC = (C + STInv) + (AR + Inv + OCA) – [(C + STInv) – WC] – (AP + OCL),

or,

NWC = (AR + Inv + OCA) – (AP + OCL) – WC.

This form for the net working capital is quite close to the form typically used in standard accounting practices. Two differences are immediately apparent. The above form does not include the short-term debt but does include the working cash.

To complete the derivation of the total tangible invested capital, we may write the net fixed asset value as follows:

NFA = TA – TCA – TIG,

where

TA is the Total Assets of the company,

TCA is the Total Current Assets of the company, and,

TIG is the total amount of intangible assets and goodwill declared by the company.

We may obtain the earnings yield, {1/E}, and the return on invested capital, ROIC, as follows:

{1/E} = EBIT* / EV,

and,

ROIC = EBIT* / (NWC + NFA).

Whilst deriving the form for the net working capital, we observed that the short-term debt was removed. Also, it is essential to recognize that the taxes due and payable for the most recent fiscal quarter for a compay must be excluded from the current liabilities. To observe the effect of these two quantities on the computed earnings yield and return on invested capital, we conducted a study in which these two quantities were either separately included or excluded from the net working capital or both were included or excluded. The results are given in the following table.

Table 1: The effect of either including the short-term debt and taxes payable in the formulation for the net working capital. The fundamental data used to compute the values contained in this table are for PWER as at 31st December 2010.
Quantity   no STD&Tax with STD with Tax with STD&Tax
{1/E}   28.55% 27.05% 28.55% 27.05%
{ROIC}   127.2% 168.8% 234.3% 429.9%

It is immediately obfious that including the taxes payable in the net working capital, most likely as a “other current liability”, has the effect of dramatically raising the value of the return on invested capital.

Let us consider two identical companies with somewhat different tax structures. Say Company A owes taxes as at the end of the fourth quarter of 2010 in the amount of 51.8695 and Company B owes 103.739. Also assume that both companies include the taxes payable as an “other current liability.” The values obtained are as follows:

Table 2: The earnings yield as a function of the taxes payable for two otherwise identical companies. Company A owes taxes in the amount of 51.8695 and Company B owes taxes in the amount of 103.739. Assume that both companies include the taxes payable as an “other current liability.” The fundamental data used to compute the values contained in this table are for PWER as at 31st December 2010.
Quantity   Company A Company B
{1/E}   28.55% 28.55%
{ROIC}   164.84% 234.31%

What we see is that by including the taxes payable in the category of “other current liabilities”, the company that is not tax efficient obtains a higher return on invested capital. Given that the mechanical screening techniques gives a one to the highest return invested capital, the best ranking, then, would go to the company that was least efficient with managing their taxes.

Let us now consider the same two companies in a scenario for which the savings on taxes by Company A was contributed to the cash holding of the company. Table 3 contains the corresponding earnings yield and return on invested capital.

Table 3: The earnings yield as a function of the taxes payable for two otherwise identical companies. Company A owes taxes in the amount of 51.8695 and Company B owes taxes in the amount of 103.739. Assume that both companies include the taxes payable as an “other current liability.” Further, assume that Company A is able to contribute 51.8695 to its cash holdings. The fundamental data used to compute the values contained in this table are for PWER as at 31st December 2010.
Quantity   Company A Company B
{1/E}   30.09% 28.55%
{ROIC}   164.84% 234.31%

The result is that the earnings yield of Company A is slightly improved, but perhaps not enough to offset the superior rank of Company B with respect to the return on invested capital.

Let us examine a scenario in which Company A invested the net tax savings in new capital assets to generate a proportionately higher income in the next fiscal quarter. We know that if neither the short-term debt nor the taxes payable are included as current liabilities when computing the net working capital, the return on invested capital is 127.15%. Thus, investment of taxes saved into new fixed assets should increase the revenue by an amount equal to 1.2715 times the amount of saved taxes. For purposes of this analysis, we ignore the taxes on the additional revenue. The results are given in Table 4.

Table 4: The earnings yield as a function of the taxes payable for two otherwise identical companies. Assume that Company A is increases its tangible working capital by 51.8695 for the next fiscal quarter. The fundamental data used to compute the values contained in this table are for PWER as at 31st December 2010.
Quantity   Company A Company B
PPE   115.195 63.325
Revenue   1904.779 1047.1
{1/E}   113.44% 28.55%
{ROIC}   411.23% 234.31%

In this simple analysis, it becomes clear that by managing the tax burden efficiently, Company A stands to see an earnings yield, all other things held the same, in the next quarter of 113.44% as compared with the earnings yield of Company B. Of greater import, Company A will be able to provide the investors with a return on invested capital in the next quarter of 411.23%, all other things held fixed, as compared with the 234.31% for Company B. By including the taxes payable in the other current liabilities, Company A obtains an inferior ranking in the mechanical screen as compared with Company B, even though Company A was more tax efficient and was able to realise a considerable increase in revenue in the ensuing quarter by investing the saved taxes.

To compare two otherwise identical companies, it is essential to exclude the current tax liability from the computations for the net working capital of the company. Since short-term debt would be handled in the same manner in computing the net working capital as was the current taxes payable, it is also essential to exclude the short-term debt. The logic behind this is as follows: consider two otherwise identical companies; Company A maintains a very low level of short-term debt, whilst Company B is not so prudent. Even though Company A carries a lower short-term debt, it will be ranked inferior to Company B if that short-term debt is included in the calculation of the net working capital. However, Company B will, most likely, suffer from two possible maladies: a portion of any revenue generated by the company will be used unproductively to pay interest and principle on the short-term debt; and, Company B is at a higher risk of defaulting on the debt and, therefore, drive shareholder value down.

When confronted with a choice between two otherwise identical companies, the prudent investor will drive down his risk of exposure to a default by choosing that company that properly manages its long- and short-term debt and operates in a tax efficient manner.

It is, therefore, possible to use a mechanical screen for efficient use of capital resources by management. Management that operates the company efficiently will seek tax efficient alternatives amongst various possible trajectories and will use only an optimum amount of short-term debt. By excluding both the short-term debt and the current taxes payable from the computation of the net working capital, the mechanical screen can separate the management of lesser quality from the superior management team.

I hope that this helps.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: