Margin of Safety

No investment is without risk, and when it comes to choosing a company to invest in, one of the most important goals is to not just do your best to avoid a loss of capital in the short term (even though this can be extremely difficult to avoid as one should always look at an investment for the medium to long term), but ensuring that in the long run you mitigate the risk of investing in companies which are at risk of experiencing serious financial difficulties and causing a permanent capital loss.

The three most common ways in which a total capital loss can come about is fraud, company bankruptcy and the manipulation of financial statistics such as the income statement, balance sheet and the amount of debt a company owes.

In many cases, the upper management of a company have been seen to be manipulating the financial statements of a company in order to make it look like it is in a much better financial position than it actually is. This is fraudulent activity and usually means there is something even more foul at play in a company such as the squandering of funds and embezzlement.

From a more straightforward point of view, even if there is a company that is involved in honest trading practices due to poor financial management and distress poses a high risk of bankruptcy is an extremely dangerous investment.  Therefore, any company which poses a loss of capital risk whether it be from a fraud or high bankruptcy risk point of view is effectively deemed “un-investable” due to the fact that the risks are simply too high and not even worth the potentially high reward.

Spotting Earnings Manipulation and Fraud

There are a few ways in which an investor can spot potential embezzlement and earnings manipulation, and this is by analyzing the scaled total accruals (STA) and the scaled net operating assets (SNOA). The first (STA) can help to spot early signs of earnings manipulation whereas the SNOA gives a snapshot at the past history of a company’s management and how they may have tried to manipulate the earnings.

Many Initial Public Offerings (IPO) companies show exceptional earnings when in fact this is due to accrual adjustments to the account. If, earnings are actually skewed by the accruals process it can cause an inflationary stock price not true to the real intrinsic value of the stock. This can result in a false sense of security resulting in a share sell off at a later stage ultimately causing a big dent in the overall stock price of a company. Typically the more aggressive and inflated the manipulation, the more a stock price can shoot up which ultimately means the correction or crash can be fast and aggressive.

Beneish M-Score

The M-Score or PROBM calculation, pioneered by Dr. Messod Beneish in 1999, is a way of assessing the chance of manipulation in any given instance. According to Beneish, it consists of the following criteria:

  • Days Sales in Receivables Index (DSRI)
 DSRI = (Net Receivablest / Salest) / (Net Receivablest-1 / Salest-1)

Any increases on a large scale between years could be an indication that management are misrepresenting and showing higher than true company revenues.

  • Gross Margin Index (GMI)
GMI = [(Salest-1 – COGSt-1) / Salest-1] / [(Salest – COGSt) / Salest]

This is the ratio between the gross margin in year t-1 to the base year, year t.

When the value of this ratio is above 1, it is a sign that the gross margin has fallen, and a firm in this situation could be more inclined to manipulate figures as a result.

  • Asset Quality Index (AQI)
AQI = [1 – (Current Assetst + PP&Et + Securitiest) / Total Assetst] / [1 – ((Current Assetst-1 + PP&Et-1 + Securitiest-1) / Total Assetst-1)]

The AQI is presented as a ratio of noncurrent assets (excluding equipment, property and plant) to the total assets a company has.

This can give an idea of how likely a company may try to defer costs through the use of intangible assets.

  • Sales Growth Index (SGI)
SGI = Salest / Salest-1

Measured as the ratio between year t sales to year t-1 sales.

The assumption here is that excessively high sales growth can create pressure within management to keep this up which can lead to a higher chance of earnings manipulation at times where sales may be going through a slowdown.

  • Depreciation Index (DEPI)
DEPI = (Depreciationt-1/ (PP&Et-1 + Depreciationt-1)) / (Depreciationt / (PP&Et + Depreciationt))

The depreciation index is measured again in a ratio form, as depreciation in the year t-1 to year t. If the index is above 1, it is an indication that any assets are being shown to be depreciating at a slower than normal rate. While this could be genuine, there is also a possibility this figure is being manipulated, as it will naturally help earnings look healthier.

  • Sales General and Administrative Expenses Index (SGAI)
SGAI = (SG&A Expenset / Salest) / (SG&A Expenset-1 / Salest-1)

This is a measure between year t and year t-1. Companies with a quickly growing SGA could indicate managers are boosting the company’s value in the form of higher employee salaries.

  • Leverage Index (LVGI)
LVGI = [(Current Liabilitiest + Total Long Term Debtt) / Total Assetst] / [(Current Liabilitiest-1 + Total Long Term Debtt-1) / Total Assetst-1]

The leverage index represents how much borrowing a company is taking on and the higher this figure is (typically greater than 1), the more likely a company is to default on a loan.  It is shown as a ratio of total debts to total assets in relation to year t to year t-1.This would put the company at a higher risk of manipulating its figures.

  • Total Accruals to Total Assets (TATA)
TATA = (Income from Continuing Operationst – Cash Flows from Operationst) / Total Assetst

Also known as the total accruals to total assets, this is calculated by taking into account the change in working capital accounts other than cash, less depreciation. The higher the accruals, the higher a chance of manipulation.

According to Dr. Beneish, the formula to calculate the M-score is:

M-score = −4.84 + 0.92 × DSRI + 0.528 × GMI + 0.404 × AQI + 0.892 × SGI + 0.115 × DEPI −0.172 × SGAI + 4.679 × TATA − 0.327 × LVGI

A resulting score of less than -1.78 suggests that a company is unlikely to be a manipulator, whereas there is a good chance a company is a manipulator if this figure is greater than -1.78.

The M-Score was tested  on The Enron Corporation and successful in its application.

Identifying Companies with a Higher Risk of Encountering Financial Difficulties

As mentioned above, not only are fraudulent companies a no go from an investment viewpoint, legitimate entities which are poorly run with a lack of focus on finances can be a high risk investment with bankruptcy a real possibility.

One of the ways used to measure the chances of a company on track to go bankrupt is the Z-Score, a formula created by Edward Altman. This formula allows for several factors such as profitability, liquidity, solvency and of course leverage.

The formula is as follows:

ALTMAN Z-SCORE = 1.2A + 1.4B + 3.3C + 0.6D + 1.0E

The nearer a company’s Z score is to 3 and above, the higher the chance a company is in a strong financial position. The closer a company’s score is to 0, the higher the chance it is likely to go bankrupt. Initially, the threshold for a company going bankrupt was 2.675 (so a company with a Z-Score below this level was considered to be heading for bankruptcy). This was then reduced to 1.81 once a number of additional tests were carried out. Therefore, it can be argued that a score of 2 or above is a “safe zone” in a sense, and represents a company with a low risk of going bankrupt.

Ohlson “O-Score”

This measure was created by James Ohlson in 1980, using four variables in order to project the likelihood of a company’s stock (and therefore the firm itself) collapsing. These four variables were the overall financial performance, the liquidity on hand, the overall financial structure and the total size of the stock.

Shumway Hazard Model

According to Charalambakis et. Al, the Shumway (2001) hazard model is a more accurate measure in comparison to both the O Score and the Altman Z score.

This hazard model takes into account accounting data as well as crucial market data. With the other two models being solely accounting focused, the argument is that this extra information on market conditions and the wider competition gives a more accurate “birds eye view” of a company’s overall situation as opposed to just the internal financial workings.

Campbell et. Al (2008)

According to their research paper “In Search of Distress Risk” in 2008, Campbell et. Al believed they could further enhance the predictability of a potential stock going into bankruptcy by building on Shumway’s model.

Rather than solely relying on accounting data like Ohlson and Altman, Campbell et.  Al consider the stock market as a whole and the according variables. In their model the inputs are as follows:

NIMTA = weighted average (net income of a particular quarter/ MTA) 
MTA = market value of total assets = book value of liabilities + market capitalization 
TLMTA = total liabilites/MTA
CASHMTA = cash and equivalents/MTA
EXRET = weighted average (log(1+stock’s return) – log (1+ S&P 500 return))
SIGMA = annualized stock’s standard deviation over the previous 3 months
RSIZE = log(stock market cap / S&P 500 total market value)
MB = MTA/adjusted book value
Adjusted book value = book value + .1 x (market cap – book value)PRICE = log (recent stock price), capped at $15

Using this methodology, Campbell et. Al use a “logit model” which gives a simple result as to whether a stock is financially distressed or not. The formula is as follows:

Pt−1(Yit = 1)  =     1
           1 + exp(−α − βxi,t−1)’

It is also important to consider whether to exclude these high risk stocks from a potential portfolio. Whilst for some investors certain high risk stocks may be worth the risk, when looking at a more stable portfolio, by removing potential stocks which flag up based on the methods above can not only reduce risk, but also look to increase overall net gains.


As we know it is imperative to carefully analyze any potential stock purchase, and by incorporating the above formulas and prediction models, an investor is able to create a large and safe margin of safety when adding to their portfolio. The main objective is firstly to identify any potentially fraudulent companies and then also look into which companies seem to trade ethically and honestly and then assess whether any of these companies are also at risk of any financial trouble. By using the methodology such as the Z Score and O Score, investors can ascertain the likelihood of a company falling into financial trouble, however these models are solely focused on the accounting principles as opposed to the market as a whole. As we have seen, building on the models by Altman, Ohlson and the Hazard Model by Shulman, investors should perhaps also consider also taking into account market variables and market conditions, and using the analysis mentioned by Campbell et. Al in 2008 be able to form a more detailed opinion on a given investment.