Sensitivity Analysis for Simulations Plus

Introduction

In the midst of grading season, I don’t have much time to blog or respond to comments, but, as I have in the past, I want to respond to a comment on a discussion board, and use this as an opportunity to do a little teaching.  The discussion board comment questioned whether Simulations Plus (a small pharmaceutical-research software company) has run up too much to still be considered a value stock.  When I first purchased it on October 21, 2008, I bought in at $1.06 and $1.07 a share and estimated that it was selling at more than a 50% discount to intrinsic value.  Today, the closing price was $2.35, and it is natural to wonder whether it is time to sell.

To answer this question, it is important to understand that my shares in SLP are not in a tax-efficient account, and, despite a market that appears to be marginally over-valued, my inclination is to hold stocks until they are demonstrably overvalued when they reside in this account.  Otherwise, the 15% capital gains tax hit reduces my proceeds on the investment by 15%.  Additionally, if selling, I would confront the dilemma of finding a replacement investment about which I am less familiar and, consequently, less comfortable.  So, if SLP is fairly-valued, I’ll hold, and, if it is overvalued, I’ll sell.

Background

Now, it should be understood that the stock hovered for sometime following purchase around the delisting price of $1 and eventually appreciated in value until hitting strong resistance around $1.80.  It meandered for a time at this level and just recently moved up strongly on increased volume, as it began to receive increased press and showed up on technical-analysis screens — where traders identify stocks that are trending higher, based on the stock-price movement and volume.  The catalyst for this appears, in part, to be attributable to the company hiring an investor relations PR firm — a move which was criticized by some on the discussion boards, with the assertion that doing so was tantamount to squandering investor value.  Personally, I take the view that you, either, trust the management or you do not, and, if you do not, then it is time to sell.  At minimum, you can buy US Government Bonds, where management can be trusted to print the money necessary to make good on its obligation to you as the bondholder.

In any event, the recent run-up was based on what seems to have been institutional buying; although, share repurchases may account for some of this move:

So, the question is one of whether recent buyers who began purchasing at an 80% premium above my entry price know a good thing when they see it or have been suckered into the stock at a fair or, for that matter, a rich price.

Weighted Cost of Capital (WACC)

To get at this, lets perform the analysis using a technique not previously covered on this blog.  Due to time constraints, I won’t explain the math, but the intrinsic value of the company can be calculated by multiplying the accumulated equity per share times (ROC-G)/(R-G), where ROC is Return on Capital, G is the growth rate, and R is the cost of capital.  To be conservative, I’ll use Cash Return on Invested Capital (CROIC) in place of ROIC (Return on Invested Capital).  CROIC is calculated as Free_Cash Flow/(Total Equity+Total Liabilities-Total Current Liabilities).

Weighted Average Cost of Capital (WACC) is normally considered the cost of capital, which I have calculated as follows:

Now, let’s look at WACC more closely.  There are two portions to this — a debt portion and an equity portion.  The debt portion takes into account the tax benefits of  debt (from the company’s perspective), but SLP has no debt and hasn’t for some time.  This means the bulk of the 13% WACC is attributable to the equity portion.

Is WACC the Best Measure?

Well, the equity portion of WACC is curious.  The company pays no dividend, so there is no cost to the company related to stock ownership when it comes to financing yearly operations.  The “Opportunity Cost” of the equity is just the investor’s required return, which, in the absence of a dividend, is the stockholder’s expected/desired price appreciation as shares are bought and sold in the market — something management can eventually influence with profits, retained earnings, growth, and increasing equity but, otherwise, management is largely powerless to guarantee a specific level of stock-price appreciation.

This 13%, however, is not a number pulled out of thin air — the air was more thick than that, even if just marginally so.  That 13% is based on the Risk Free Rate of Return (RFRR), which is the current yield on the 10-year bond, plus a buffer that accommodates market risk (8.6% is the number I use, and, by most accounts, it is indefensibly high).

Indeed, for a company with no debt and a nice cash stockpile, 13% or more strikes me as too high, and, for companies with a much weaker balance sheet, 13% strikes me as too low.  Given SLP’s strength, it seems reasonable to substitute the 20-year Single-A corporate bond yield of 6.25%.  Why?  Well, with no debt, ownership of the company is uncontested beyond just us stockholders, in that no creditors, preferred stockholders, and warrant-holders enjoy a senior claim on the company, and this makes the corporate bond yield a more reasonable measure of the true cost of equity capital — especially, since I am not using the lower rates (yields) associated with Triple-A bonds.  This seems (to me, at least) to represent a better balance between the opportunity cost of equity and the negligible effect of equity on the cost-burden faced by the company in its daily operations.

Would Warren Approve?

Now, you may assert that Buffett famously requires a 15% rate, but Mr. Buffett is notorious for allowing common sense to trump general rules of thumb, even when those rules are of his making.  In other words, he does not check his brain at the office door upon arrival each morning.  By his own admission, the purchase of Coke was made at a fair price (rather than a DCF-based cheap price) — allowing the franchise value of the company to serve as his margin of safety.  In the case of SLP, the company is the single producer of its leading product and the government, in addition to many leading pharmaceutical firms, is literally sold on it.  So, I’ll use 6.25% as the cost of capital, and, if you (dear reader) disagree, feel free to substitute any number you prefer.  In fact, that is one of the strengths of performing sensitivity analysis, as will soon be evident.  Sensitivity analysis will allow you to select your own inputs and quickly identify your own estimate of intrinsic value, based on your preferences.  Having settled on a cost of capital figure, we will, however, need to identify ranges for Growth and CROIC.

Returns on Capital and Growth Rates

As for which numbers we will use for return on capital and the growth rate, we want to lower both since this method of estimating intrinsic value seeks to identify the long-term rates.  In the case of growth, we will start with the average rate of inflation over the past decade (3%) and increase that up to Mr. Buffett’s maximum expected rate for the broader-market (6%).  While he hasn’t fully explained why he believes this is the maximum for the broader market, he has implied that this has to do with the US’s current account deficit, and, indeed, some economists expect inflation to rise to 5% due to increased government spending.  If expecting that SLP will earn returns that do nothing more than keep pace with inflation, we can use sensitivity analysis to calculate intrinsic value at different rates of growth, and our span of considered rates will run between 3% and 6%.

As for the return on capital, the yearly figures over the past decade are:

If we calculate the median of the rolling five-year and seven-year medians for CROIC, we get:

So, we will adjust off of 15.59%, and our adjustment will be downward.

This raises the question of just how far down we should adjust.  We know that, over time, the returns on capital will move down to meet the weighted average cost of capital, since the spread between them will attract competition.  While a price war may lower it further, a negative spread will promote industry consolidation in order to increase pricing power and a return to reasonable profitability.  While WACC was not appropriate when calculating the cost of capital (because investor expectations are not relevant to how the company operates daily), WACC is the appropriate measure for estimating the low-end of long-term returns on capital — because rates lower than that will bring on an investor up-rising (in the form of proxy fights, the arrival of activist investors, etc.).  Consequently, we will use 10% as our bottom returns on capital estimate and 15% as the top-end estimate.

Sensitivity Table/Analysis

The final preliminary step is to plug each of these assumptions into our equation for calculating intrinsic value.  Using 6.25% as the cost of capital, our two-way sensitivity table for growth and returns on capital becomes:

So, at 4% growth and 13% returns on capital, the stock is worth $2.49, which is a little above today’s price.  This suggests that recent buyers were not entirely delusional.  They may, in fact, expect a higher rate of inflation-driven growth, which would value the stock at $4.49 if inflation rises to 5% and long-term returns on capital decline to 13%.  With the company’s strong niche servicing the cost-reduction needs of the pharmaceutical industry (in this post-healthcare-reform environment), higher rates for either growth or returns may be warranted, depending on investor opinion about the company, its products, and their prospects in the competitive market.  A more important consideration for the institutional investor is calculating the down-side risk in comparison to the potential for appreciation.  Worst case (using 10% returns and 3% growth) is a price of $1.34 … unless expecting something worse to unfold over the life of the company.

Now, the table above color codes the cells based on whether the projected value is above (green) or below (red) the current stock price.

Down-Side Risk Graph

Raw numbers, however, are often difficult to visualize.  So, think about the graphing possibilities.

Want the ability to consider the down-side risk versus the up-side potential in light of these assumption?  How about:

In this case, the stock prices are listed along the y-axis, returns on capital are along the x-axis, and the diagonal colored lines are the various growth rates.  There are three horizontal lines.  The red-dashed horizontal line is my test assumption, which assumes 14% returns and 5.5% growth.  Do I expect this?  No.  It is just an example of what you can do with this sort of analysis.  The dashed-blue line is set at twice the current stock price.  As a deep-value investor, I tend to demand a 50% margin of safety when first buying a stock, and this line helps me determine the scenarios necessary to produce that margin of safety.  At 11% returns and 5.5% growth, intrinsic value doubles — both strike me as achievable, for the reasons explained earlier, but you may view things differently and this chart informs that consideration.  Lastly, for this chart, the black line is the current price, and, to the degree that the different growth rate lines are below the black line (compared to the degree to which they are above the black line) provides a sense of the down-side risk.  In this case, the upside seems abundantly larger than the down-side, but you may be exceedingly risk-adverse and see it differently … and that is fine.

Jacobi — Charlie Munger’s Hero, and Mine, As Well

Continuing with our graphical analysis, let’s do the Jacobi and invert:

In this case, the price remains along the y-axis, but the growth rate is along the x-axis.  Each of the variously-colored curved lines is the return on capital, and horizontal lines reflect the current price, double the current price, and the estimate — from bottom horizontal line upwards.  Note where the top and bottom lines cross the current price and the degree to which they are above and below it.  Next, do the same with the margin-of-safety price.  And, finally, note the change in slope as growth and returns on capital increase.

With some stocks, the slope becomes parabolic, asymptotic, and down-right exotic … and you never thought a graph could achieve such an alluring level of  sex appeal.  Well, if the graph seems too good to be true, consider the before and after pictures of May West, who was considered the sexiest woman of her era when she was young and became a real porker when the ravages of age struck.  The same is often true of fast-growing companies.  Attractive in youth, they tend to burn-out early through hard-living and, like Mae, unsustainable growth.  [Note:  The general rule is that growth in excess of 30% per year threatens detrimental burnout, but, as always, there are exceptions to this rule.]

Do I Need More Analysis or Am I Just Sensitive?

Another way to analyze this is the following:

Like the earlier table, this one is color coded.  It considers the projected intrinsic-value price in comparison to the current price and calculates the up-side return.  “Fair value” is the current price, and fair value equals 100%.  Below 100% is color coded as red.  The yellow squares range between 100% and 150%, and they are yellow because I consider this to represent an attractive valuation for only exceptional companies — where management walks on water, divides loaves and fishes, births babies, washes windows, and sheers sheep before the rest of us have had breakfast.  If Warren were younger or immortal, I’d buy Berkshire in the yellow zone.  The green boxes (without bolding and lines drawn around them) indicate an intrinsic value that is between 150% and 200% above the current stock price.  This would apply to much of the Fortune 100 in terms of their size, financial strength, market share, etc.  This tier is below my normal deep-discount threshold, and examples of recent purchases that fell into this range are Disney and Wal*Mart.  Lastly, the bolded and boxed-in green squares are 200% or above the current value.

A Final, But Important, Note

For the record, I tend to calculate these figures using several different methods of estimating intrinsic value.  Think of it as Sensitivity-Analysis-Squared.  Here is one of many possible examples, using the SLP data:

Credits and References

Finally, for a more complete explanation of the math behind the equation, read Value Investing:  From Graham to Buffett and Beyond by Bruce Greenwald and Judd Kahn, Paul Sonkin, and Michael van Biema.

The charts and graphs do not appear in the Greenwald book.  For good or ill, they represent my effort to incrementally improve on Professor Greenwald’s brilliance.

The purpose of this posting, beyond responding to an online comment, was to teach the concept of sensitivity analysis and take it a step beyond earlier efforts that focused on the Ponzio approach to Discounted Cash Flow analysis:

Example of Sensitivity Analysis Using Ponzio DCF Approach and SLP

The original idea for addressing this topic followed from a recent conversation with Bill Ellard.  Bill has written and published for a number of online investing Zines, while the Ponzio approach can be found at FWallStreet.com, where professional money manager Joe Ponzio is the chief cook and bottle-washer.  [Joe also runs Ponzio Capital and is the author of F Wall Street.]

Disclaimer

Now for the disclaimers, which have not been approved by any competent legal mind.  First, be an adult and do your own analysis and due diligence before investing.  I am neither perfect nor clairvoyant when it comes to investing, and I’ve made my fair share of errors in the past.   While not intending to fool anyone, I have made mathematical mistakes before (starting in first grade and continuing uninterrupted to the present day).  Consequently, there is no guarantee that the math presented here is perfect (I’ve checked it thrice or more, but I am, both, writer and editor, and, therefore, violate the tenet that no physician should have himself as a patient).  So, do your own thinking and ciphering, make your own decisions, and be adult enough to withstand the consequences of a mercurial market, an imperfect writer, and the potential that the financial Gods just might smote us both, no matter how good looking or worthy of wealth the two of us might be.  Please know that it is my preference that you not invest in any stock used as an example on this blog.  Instead, use the concepts to identify your own investment opportunities and send me a note with your best ideas … but only the ones that make a change of pants or panties necessary … and quantitatively explain your logic.  [Note: Good or bad, I will not post your recommendations.  Sorry, I shill for no one.]

_________________

Following publication of this piece, Mike wrote:

Thanks for a great report regarding Simulation Plus,Inc.. You inspired me to due my own due diligence in this company. I am new to investing but here are my results:
ROE over the past three yrs., 8.8%,14.3% & 14.5% ave. 12.5% Would like to see 15% or more;
Growth Rate for yrs 05/06 23.3%, 06/07 51.1%, 07/08 1.2%, and 08/09 1.89% Big drop here could be trouble.
Net Profit Margin is on ave., 15.6% not bad;
Asset Turn 0.76 cents,0,86 cents and 0.75 cents not good;
ROA last three yr. ave is 12.5 % could be better;
FCF 06/ -0.22,07/ +2.33,08/ +1.67 and 09/ +1.77 this seems lumpy;
Cash King Ratio for four yrs. is on ave., 15.11% which is very good;
Cash Conversion Cycle for Simulations is 116.7 days I don’t know if this is good or bad?
Because I am new at this I need your ideas if you still think SLP is still a good investment. Thank you for your time and consideration in this matter.
Respectfully, Mike

First, Mike, thank you for the compliment.  They are always appreciated.

Second, please recognize that I used SLP as an example to demonstrate an analytical approach.  While I own shares in the stock, SLP is a very small holding in a larger portfolio, and this strikes me as appropriate for a small value/growth company with a market cap significantly less than $100 million.  During a significant market down-turn, market liquidity could dry up quickly.  So, while the company is abundantly healthy (no debt, strong cash flows, etc.), don’t deploy your rent money or allocate Granny’s savings toward it.

Morningstar Data.  See Debt/Equity figures.

Third, your calculations differ from the Morningstar figures (marginally, in some cases, more significantly in others):

Now, let’s look at each of the items you mentioned, starting with Net Profit Margin.  At 15%, this is, indeed, healthy.  Next, you note that Asset Turnover is “not good.”  The reason we consider Asset Turnover is that it contributes to Returns on Assets when multiplied by the Profit Margin.  This tells us a great deal about the company’s pricing structure if you think of Asset Turnover as a surrogate for the volume of unit sales.  With a high profit margin product, you can sell a small number of units to achieve the same ROA as a low margin product for which you must sell a great many units.  This is the difference between Porche and bubble gum — one has high margins and low unit sales, while the other has low margins and high unit sales.  This, in fact, is recognized by Investopedia (the online dictionary of investment terms and concepts):

http://www.investopedia.com/terms/a/assetturnover.asp

Now, you note that ROA has a three year average of 12.5% (indicating that this “could be better.”)  Recognize that SLP is an early-life-cycle company in its formative years.  Consequently, the numerator (net income) will be depressed due to growth CapEx spending (plowing back into the company a greater portion of operating income to finance future returns), and the denominator of ROA (Total Assets) may be higher-priced (listed as more valuable or expensive) than for a longer-tenured firm.  For example, there is a great deal of uncertainty about the real estate value of Sears Holdings because much of the property was purchased more than 40 years ago, at much lower prices.  Credit Suisse values the real estate holdings at $4.7 billion today, while Sears lists them at $1 billion on the balance sheet.

Next, you note that the three year average ROE is 12.5% (the same as for ROA), and that you would like to see something greater than 15% by this measure.  The difference between ROA and ROE is literally the multiplier effect of debt:

http://www.investopedia.com/terms/e/equitymultiplier.asp

Recall from earlier that SLP has no debt.  Of course, debt can be a good thing or it can be a bad thing.  It is known as leverage for a reason.  It enhances profits and exacerbates losses.  Currently, corporations are stockpiling cash due to economic and tax-rate uncertainties, allowing them to reduce debt, and this is depressing ROE.  So, in this respect, SLP is not alone.  In fact, the company has indicated its interest in making strategic aquisitions, if the right deal comes along, and this explains why cash and short-term equivalents account for nearly 60% of total assets:

More importantly, the company should not take on debt if unable to convert that capital into new sales.  Doing so would depress ROA and ROE, and it would certainly depress any right-thinking investor.

The decline in growth rate is easily explained by the decline in the economy — the worse recession/depression since 1929 — and the uncertainties associated with healthcare reform legislation for SLP’s principle customers (the pharmaceutical and bio-med industries).  The question then becomes whether this represents a short-term condition or a long-term undermining of the company’s strategic position.  While this is largely a judgment call, note that SLP’s principle product significantly reduces R&D costs for its clients and the utility of the product is recognized by the FDA as a suitable surrogate for those R&D steps.  So, the question becomes one of whether you expect pharma and bio-med to cease R&D toward new medical advances — if so, be gentle when you break the news to their stockholders.

Next, you indicate that FCF is lumpy, and it certainly is.  Dive into the statement of cash flows, however, and the reason becomes fairly evident.  Note that the numbers are exceedingly small.  A change of $1 million in cash from investments can have a material impact on FCF, even though that $1 million expense was toward upgrading operating infrastructure, as happened a couple of years ago.  That decline in FCF, however, represented an internal investment toward future growth, for which the company and its owners should reasonably expect a future return on investment.  And, indeed, the new product announcements since then support this view — with one product already generating profits and the other new product just starting to enter the market, following the recently announced sales and distribution partnership agreement.

You note that the Cash King ratio (15.11%) is “very good.”  Candidly, I’m not familiar with this ratio.

As for whether SLP remains a value investment, I am of two minds about that.  At the current price it is fairly valued if expecting 6% annualized growth.  If, however, you anticipate a higher growth rate due to the company’s proven history of innovation, new products (generated internally), and the company’s past history of accretively successful and judicious acquisitions, then the stock is undervalued by 50% or more.  To see this most clearly, look at the Returns on Invested Capital and compare that to any reasonable measure of the company’s cost of capital.  The difference between the two reflects the franchise value of the company (what Buffett calls “The Moat”).

Now, I should offer two warnings to the approach you are using.

First, allow the ratios to help paint a picture of the company as a strategic entity in the market.  In other words, understand the strategic position of the company and its forward prospects and let the ratios support or disprove the mental picture you have of the company.  Because the ratios are backwards-looking (reflecting prior results), they will tell you about how management has positioned the company and whether the company is in financial distress, but they are often poor indicators of future performance.  This is especially true for early-stage firms.

During grad school, we did a case analysis of American Greetings.  The financial results and ratios were little different from one year to the next, and my team’s presentation of the case reached that conclusion.  Our professor (Dr. Doward Dowsma) then asked the key question — “Is American Greetings a stock or is it a bond?”  Well, of course, it is a stock, but its behavior was more akin to a bond — with a steady and predictable revenue stream and dividends.  Dr. Dow then asked whether it was easy or hard to value the company, and, because it behaves like a bond, we said it was easy.  “As investors,” Dr. Dow continued, “is there anything about the stock that you know that other investors do not, since the value is so easily determined?”  We could think of none.  “Is the stock fairly valued today?”  It was, we asserted.  “So, the market knows how to read a set of financials, calculate the ratios, discount the cash flows, and place a fair value on the stock?”  Yes.  “Is American Greetings a buy, sell, or hold?”  We said “hold.”  “Oh, really?” responded Dr. Dow.  “You can’t find a better use for your money than purchasing a fairly valued bond?”

Now, ask yourself, “Is SLP so predictable in its results that it behaves like a bond?”  Then ask, “Do you know something the market does not?”  Personally, I know that the company enjoys significant barriers to entry, that management is earnest and honest, that it occupies an industry that is more resilient to market down-turns than most, has a new product that may revolutionize workplace efficiency for disabled workers, and has consistently posted returns on invested capital that exceed the imputed equity cost of capital (maximally, 8.6%).  I also know that it is a small company that is under-followed by Wall Street analysts.  So, yes, I know something the market doesn’t … and I know the company is so small that it represents a speculative investment.  Consequently, I want exposure to the company, but I don’t want a lot of exposure.  That was my thinking when I bought the stock when it was selling near its delisting price of $1, and that is my thinking today at $2.35.  The difference between then and now is that the company was cheap on a book value basis at $1.

Second, if relying so strongly on ratios and conventional measures for them, you need to consider more than three years worth of data.  Three years is statistically meaningless and, worse, misleading.  Under the central limit theorem, you need 30 data points to determine when a single data point is statistically outside the normal distribution (i.e., to reject the null hypothesis).  Walter Shewhart, in the mid-1920s, demonstrated that you could use as little as five data points if the data was accurately collected and reflected a standardized process.  Well, the distribution of economic results is not captured by 10 years worth of data (the prior decade’s data was unhelpful in predicting the crash of 2007/2008), and it is certainly not captured with 5 years worth of data, and 3 years data is 40% less data than five years data.

This is why Benjamin Graham advocated requiring a significant margin of safety.  That margin of safety can come in different forms.  It can be based on a discount to the company’s balance sheet (Graham’s approach), but this entirely ignores the compounding benefits of future growth.  You can use John Burr William’s DCF to get at the value of growth, but this requires making judgments about growth and discount rates.  You can discount DCF (a la Joe Ponzio and Monish Pabrai), but this provides little protection during economic crashes that exceed 30% (or less, depending on your growth and discount assumptions).  You can do as Buffett does and combine Philip Fisher with Graham and Williams and add a heavy dose of Michael Porter’s “Strategic Forces.”  Or you can modify Buffett with Bruce Greenwald’s methods — adding layer upon layer of safety margins, where the value of each margin of safety is accorded its appropriate weight depending on the strengths of the company — something I’ve barely touched on in this blog, but I’m working on it with the latest series of quality improvement postings (but that won’t be evident until the end, due to the complexity of the theory and philosophy behind it).

In any event, I hope this is helpful.

The key takeaway is not that SLP is or is not a stock you should buy, however.  It is right for my portfolio, because I can accommodate a significant loss from a small and speculative investment, and, having experienced and navigated the crash of 2007/2008, I know with relative precision my tolerance for risk.  As a new investor, your tolerance may be lower (most investors believe their tolerance is high, until the market hits them hard) or your invested capital may be less discretionary.  So, please do not read this as a recommendation to purchase SLP.  That is not my intent, and it was not why I used it as a case study originally.  In fact, I chose SLP as the example because I believed most investors would elect against buying, due to the small market cap, low share volumes, early-life-cycle stage, etc.

Thanks,

Robert