Psychemedics (PMD): A High Quality, Low Growth Business with a Dividend Yield Over 7% – And A Third of the Business About to Disappear
Psychemedics (PMD) is a micro-cap stock (market cap around $50 million right now) that trades on the NASDAQ. It is not – by my usual definitions – a particularly overlooked stock. The beta is about 0.53 (low, but many stocks I’ve written about have much lower betas). The share turnover rate – taking recent volume in the stock and multiplying it to see how much of a company’s total shares outstanding would turn over in a given year at this recent rate of trading – is 102%. Again, that’s not an especially high number for the market overall. It might not be high compared to the stocks in most people’s portfolios. But, it’s very high for any sort of truly “overlooked” stock. There are some possible explanations here. Some recent events have caused Psychemedics stock to be especially volatile (this suggests the low-ish beta of 0.53 is more the result of very low correlation with the overall market than of actual low volatility in the stock) and insiders own very few shares of this company. In fact, there’s a ton of float and very few long-term investors in this stock. To give you some idea: insiders all own about 3% or less of the stock individually (and less than 10% taken all together) and Renaissance Technologies (a quant fund) was the largest holder of the stock as of the last proxy statement. This points to a stock that is not closely held by insiders, not generally held by long-term investors, etc. So, it trades a lot. It may not be overlooked. But, it’s still a micro-cap stock. And there’s a decent chance you haven’t seen a longer write-up of this particular stock. So, I do consider it overlooked enough to at least do an “initial interest post”. This is that post.
Psychemedics is in one line of business: hair testing for drugs. Actual testing services make up over 90% of revenue in any given year. The rest of revenue is all very closely related revenue such as charging for the actual collection and shipping of hair samples (in some cases). This is really just a drug testing company that uses hair to do the tests. That’s all you need to know. Historically, almost all revenue and earnings and so on came from the United States. Almost all costs still do come from the U.S. The company is headquartered in Massachusetts and rents about 4,000 square feet there, the actual lab is a rented facility in Culver City, California. Although the lab facility is rented – the equipment is owned. Psychemedics in an asset light business with one exception: it owns a lot of lab equipment. The company depreciates the equipment over 5-7 years. If you do the math on what the equipment was valued at gross, what cap-ex on the equipment is, etc. – this is the one asset heavy part of the business. The company’s assets really consist of technical know-how and a bunch of lab equipment for drug testing.
A few years ago – Psychemedics expanded into the Brazilian market to do testing of professional truck drivers there. I expect this business may disappear entirely in the next year or so. I mean Psychemedics may lose the business. And I am going to analyze Psychemedics as if it has already lost the Brazilian business. The market for drug testing via hair will continue in Brazil. It’ll just be dominated by someone else. This is a good place to pause and explain what drug testing using hair is.
Most drug testing around the world – and that includes in the U.S., which does also test truck drivers (like Brazil does, but using urine instead) – is done using urine testing instead of hair testing. Why? Is urine testing a better method of detecting drugs? No. It’s worse. It’s a lot worse. Psychemedics goes into some details about how much more effective hair testing for drugs is than urine testing for drugs. But, the truth is that – before opening Psychemedics’ 10-K – I was aware of a lot of this data. It’s relatively easy for employees, job applicants, students, criminals, prisoners, etc. to trick a urine test. The test only detects drug use in the last few days tops. So, the easiest way is to simply abstain from drugs immediately before you expect to be tested. It is also relatively easy to get false positives from some other things – for example, eating poppy bagels could cause a false positive on an opium test – that the threshold amounts used to define failing a urine test sometimes have to be set high (to eliminate these possible false positives). Then, there are more active methods of avoiding a positive on a urine test such as consuming abnormally high quantities of water – this combined with the need to set a higher threshold to define failing a drug test could be enough to pass some drug tests even with recent drug use – on to much more involved processes like obtaining clean urine. Obviously, it’s possible to combat many of these tricks used to avoid drug detection. Surprise drug testing would help. Using an observer to witness the actual urine collection process would help. And so on. However, many of these methods have downsides. If what an employer really wanted was to be 100% sure they had detected as many drug users as possible – they’d simply use a hair test. But, most don’t. Why not?
One possible theory – and the one I personally lean towards – is that not all entities that perform drug tests are really that eager to: 1) Pay more for hair testing than urine esting AND 2) Actually detect the maximum number of true drug users. Psychemedics gives the example of a prison where urine testing gave a 0% positive rate and then hair testing gave about a 7% positive rate. Measures were then put in place to reduce drug use – that’s a lot easier to do in a prison than with employees – and the rate was reduced to closer to 2% with further hair testing. Urine testing showed 0% positive both before and after the drug reduction methods were used. I think this – coupled with higher pricing on hair tests – is a pretty good example of what an employer or other entity using drug tests might not be that excited to find. Some organizations may want to test for drugs, say they test for drugs, make it clear they intend their workplace to be a drug free environment etc. – but, then, not actually eliminate 7% of their workforce, potential job applicants, etc. I think this is supported by the research Psychemedics includes on what HR executives say they are trying to eliminate with drug testing. The number one thing they care most about is people missing work. They believe that employees on drugs are more likely to miss work. That’s probably true. But, I’m not sure it’s necessary to do accurate drug testing to lower a lot of job absences. HR may be quite capable of screening out the less well functioning drug users from those who are quite capable of attending work regularly despite their drug use. In some industries, there are other concerns about safety and trustworthiness and so on. Anecdotally, from knowing managers who have supervised employees with drug problems – this fits. I know of a couple cases where employers knew they were retaining employees with serious drug problems and yet did kept the employee on indefinitely. In these cases, the desire was to avoid union problems, litigation, etc. In those cases, the employer specifically instructed the employee’s supervisor to use special procedures to make sure the drug user did not have access to things like cash handling that others in the position might normally have. I’m relying on that HR executive survey, anecdotes, etc. here to make an important point. If you’re reading this and you know urine testing is common and hair testing isn’t common – you’re probably thinking Psychemedics is exaggerating the effectiveness of hair testing versus urine testing.
I’m not convinced they are.
In some cases where establishing a clear history of past drug use is critical – such as in a custody dispute – the very same agency that might rely on urine testing normally for its own workforce will use hair testing instead. As Psychemedics explains, hair testing can give you a good 90-day record of drug use including more information on the frequency and severity of the drug use. There are cases where urine testing is clearly better, of course. The moment after an accident – a urine test can detect drugs. Hair testing can’t detect drug use in the very recent past – like this past week – because you’re collecting a hair sample from above the scalp. Hair growth starts under the scalp. But, most employers don’t need to know if someone used drugs within a week of their interview, start date, etc. They just need to know how often and how heavily they used drugs in the last 3 months. I think Psychemedics is right and hair testing is the best way of knowing that. But, I don’t think most companies using drug tests need that kind of information at any cost. There may be some very special positions involving safety, risk of liability coming back to the company, certain risks involving children, etc. where an employer would care so much about quality and so little about price that they’d use hair testing exclusively. But, I’m convinced the market for hair testing will always be small versus urine testing. I think sensitivity to price here is pretty high. And I think very few employers would prioritize really needing to be 100% sure about someone’s drug use.
So, Psychemedics is a niche business. It seems to have a good technology. Other companies – especially bigger labs as part of their services – do offer hair tests. But, these competitors rely on other people’s tests. They don’t design the testing methods themselves like Psychemedics does. R&D spending at Psychemedics is not high. It’s over $1 million but less than $2 million in any year. There are 6 R&D employees. That’s not big in absolute terms. Nor is it big relative to Psychemedics’ own revenue line. It’s just a few percent of sales. The company has continued to develop additional tests, get additional patents, etc. I don’t want to say Psychemedics doesn’t innovate. But, it doesn’t spend a ton on research. If you look at salary and other information – it’s clear that the head of laboratory operations is the second most important executive (behind the Chairman & CEO). The Chairman/CEO and the head of the laboratory have been with the company a long time. What you see in Psychemedics’ past history is the record built by the current management team.
It’s a high quality, low growth record. Over the last 20 years – and including the expansion into Brazil – Psychemedics has only managed to grow revenue by about 4% a year. Some of this may be due to difficulty raising prices (because this would hurt volumes versus the cheaper urine testing method). I’m giving you per share numbers. If I gave you pure companywide numbers, it wouldn’t change much. In recent years, there has been some share dilution causing a drag of maybe 0.5% a year on shareholder returns. The company doesn’t buy back stock. What it does is pay a big dividend.
Psychemedics basically pays out 100% of its owner earnings (free cash flow) in dividends. You’ll notice today’s dividend yield is 7.8% despite a P/E of 18. Those two don’t sound like they could both be correct. But, they roughly are. If I calculate out my own guess of recent “normal” free cash flow at Psychemedics, I get a payout ratio in the 80-85% range. The company has been paying out more than 80% but less than 100% of its cash earnings. It might be trading as cheaply as about 10 times its recently normal free cash flow. Psychemedics has net cash – not net debt. So, it sounds cheap. But….
Brazil is now one-third of total sales. The company pays a very high tax rate in Brazil – this is because the tax is based on sales not income and Psychemedics has very fat profit margins, so it seems like the company is paying around 50% of Brazilian income in taxes. All of Psychemedics business in Brazil has been done through a single distributor. That distributor had an agreement to exclusively market Psychemedics test in the country. This was an attractive deal, because Brazil recently added a rule that truck drivers in the country had to submit to drug testing via hair quite frequently. Psychemedics put a lot into this business. And it paid off. But, now the company’s distributor has sold more than 50% of its shares to a competitor of Psychemedics. The updated agreement between Psychemedics and the distributor is non-exclusive. The distributor can now sell its parent company’s drug tests as well. It likely will. Psychemedics warns that the competitor may not have the scale immediately to handle this volume of business. So, Psychemedics may retain some Brazil business for a little while. Longer-term, my expectation would be that 100% of this distributor’s purchases eventually shift away from Psychemedics and Psychemedics tries – but, who knows if they’ll succeed – to get another distributor in Brazil, enter the country directly, etc. Direct entrance would be out of character for Psychemedics. This company is generally a lab that performs drug tests. It does sell direct to parents online. But, that’s not a big part of its business. Marketing consists more of signing up big organizations and then collecting revenue on a price per test performed basis. So, it makes money based on the volume of drug testing of new applicants, employees being tested, etc.
Psychemedics has minimal costs associated with Brazil. It has almost no assets in the country and lists just one employee as being dedicated to Brazil. The company did pay a board member about $150,000 a year to handle corporate governance and other issues in Brazil. I suspect the director’s job was mainly to handle relations with the distributor and secondarily – perhaps – to avoid things like any foreign corrupt practices act (U.S. bribery law) violations in the country.
So, if we plug in a 35% to – because there will be negative impact due to operating leverage going the wrong way – like a 50% decline in normal earnings, FCF, etc., we’d be talking about paying more like 15-20 times future FCF (without Brazil) for the stock. That 4% historical growth rate was achieved including the new Brazil business.
That leaves a free cash flow yield of maybe 5-7% a year or something on today’s price. Growth might be 4% or less. We’re getting numbers that don’t add up to much above 10% a year. And then this isn’t a “Davis Double Play”. The P/E is already in the teens. I’m saying the Price/FCF might be 15-20 times future FCF. We shouldn’t expect multiple expansion here.
If all goes badly in Brazil, this looks like it might be fairly priced – not much underpriced. It’s a high quality business. I think it’s likely – but not certain – to outperform the S&P 500. Psychemedics has paid a dividend for over 23 years now. The dividend yield is over 7%. Even after losing Brazil, this company can – and almost certainly will – pay big dividends in the future.
This isn’t a stock where you are at risk of losing money, getting low single digit returns, etc. But, so far, I don’t see a very certain path to returns over 10% a year.
For now, I think the stock is probably a pass. But, it’s one I’ll keep a file on for future investigation.
Geoff’s Initial Interest: 80%