I did not invest in J.C. Penney (JCP, Financial). I was severely tempted, several times, because it seemed like a good company to coat-tail. It felt like a stock with small downside and arguably huge upside. The newly hired CEO (Ron Johnson) gave up $100 million in stock options at Apple (AAPL, Financial) and put down $50 million of his own money to buy stock of JCP. The incentives were aligned. Johnsons’ track record in retailing was also fabulous. He had worked at Mervyns, Target (TGT, Financial) and was the pioneer of the Apple Retail Stores. JCP seemed like a good asymmetric bet. Given its extensive real estate ownership and a well recognized brand name, it seemed that given time, Johnson will be able to turn the company around.
If I had taken the time to read the investment thesis of JCP, read their annual reports and make my own due diligence, I would probably have bought the stock. As it is, I had to control myself several times from pulling the trigger. I succeeded only because I felt that JCP was a high-risk scenario, and I have better things to invest in, like Bouygues (XPAR:EN) and CAF (XMCE:CAF).
Reading the article by Vitaly Katsenelson (What I Learned from the JC Penney Fiasco), I tripped on the following assertion.
If Vitaly’s assumptions are correct, i.e. JCP had a 70% likelihood of 300% return and 30% possibility of -40% return, then the expected return for this endeavor was 198%.
The determination of expected return hinges on three important variables: the probability of success, the upside and the downside. The upside is much easier to deal with. One can use several valuation tools to calculate the “normalized” value of the company and hence the upside.
The real estate value of JCP was the downside protection in this case. Taking the real estate value and subtracting the net debt of the company was not a good way to measure the downside of the company.
I will momentarily defer the discussion on calculating the downside and talk about the third variable in the calculation of the expected return, i.e. the likelihood of success. This is probably the most important variable in the discussion, and we must at least make sure that our process for arriving at this value is correct.
The task seems intractable. Vitaly comes up with a 70% chance of success. But what is the process behind the number? Is it just a number which he thought plausible? Or was there a process involved?
The ever quotable Mark Twain has a solution hidden inside his famous quote. “History does not repeat itself, but it does rhyme.” The answer is to look at the history of such turnarounds. As much as we would like to believe that we are unique, someone has gone through very similar experiences and has fantastic lessons for us.
One might start by answering the following questions.
How many retail turnarounds have been successful? You might try to find cases which bear similarity to the case in hand like a star CEO being hired to save an ever-fading brand. If you do not have data on retail turnarounds you can instead look for the cases were hiring a good CEO to save the business has worked (or not). This will give you a baseline estimate for the success.
What was the time it took them to become successful? This will give you an estimate on the number of years you have to hang around for the company to realize its value.
What is the track record of the CEO, if any? Always adjust for luck in this case. If the CEO has been successful once or twice, the chances are higher that he will fail this time. The luck is bound to turn at some point.
Once we have a baseline probability of success, we can adjust it according to the situation we find ourselves in. Buffett has mentioned that “retail turnarounds rarely turn” and I venture to guess that the 70% probability of success was quite optimistic. It means that hiring a very good CEO will lead to turnaround in 70% of cases.
I think a better proxy will be closer to 30%. I emphasize that I am pulling this number out from thin air and probably committing the same mistake. But this number seems more plausible because the process to reach it is less arbitrary.
Coming back to the downside, a very similar process is advisable. You start with the average duration of the turnaround and then come up with the losses/debt taken in the intervening period. This will lead to a smaller value than just taking the real estate value and subtracting the net debt.
I trust Vitaly on the calculation of the downside. I respect his opinions and have learned from his writings.
If we run the calculations for the expected return with 30% chance of success with 300% upside and 70% chances of failure with 40% downside, then the expected return reduces to 62%. If we assume that it would have taken JCP at least two years to turn itself around, a very conservative estimate in my opinion, then the compounded return reduces to 27% a year.
The investment is not so attractive anymore.
If I had taken the time to read the investment thesis of JCP, read their annual reports and make my own due diligence, I would probably have bought the stock. As it is, I had to control myself several times from pulling the trigger. I succeeded only because I felt that JCP was a high-risk scenario, and I have better things to invest in, like Bouygues (XPAR:EN) and CAF (XMCE:CAF).
Reading the article by Vitaly Katsenelson (What I Learned from the JC Penney Fiasco), I tripped on the following assertion.
“We thought there was a very high likelihood — 70 percent or so — that Johnson would be successful; if so, the upside would be threefold or more. If he failed — a 30 percent chance — the downside was probably 40 percent or so. The risk-reward scenario was very attractive.”It led me to think about the following question: How does one go about calculating the probabilities of investment success versus investment failure? It is a very important question because I completely agree with Vitaly when he says that “uncertainty is the nature of investing.”
If Vitaly’s assumptions are correct, i.e. JCP had a 70% likelihood of 300% return and 30% possibility of -40% return, then the expected return for this endeavor was 198%.
The determination of expected return hinges on three important variables: the probability of success, the upside and the downside. The upside is much easier to deal with. One can use several valuation tools to calculate the “normalized” value of the company and hence the upside.
“If Penney achieves the pre-Johnson level of $150 per square foot and gets to keep $700 million of cost cuts, its earnings power will be $3 to $4 per share. If sales per square foot come back to the 2007 peak of $170, earnings will jump to $6 a share.”Ever optimistic, the downside is generally ignored. In many cases, investors plug in an arbitrary value which seemingly looks reasonable. But it is important to come up with the process to calculate the downside. The situation of JCP is a familiar one. When the new CEO is hired, he probably has enough money and good enough balance sheet to turn the company around. But the company starts bleeding money, it needs to take on additional financing and the book value deteriorates as time passes. Glancing at Morningstar, the book value of JCP has gone from $20 in 2010 to $7.
The real estate value of JCP was the downside protection in this case. Taking the real estate value and subtracting the net debt of the company was not a good way to measure the downside of the company.
I will momentarily defer the discussion on calculating the downside and talk about the third variable in the calculation of the expected return, i.e. the likelihood of success. This is probably the most important variable in the discussion, and we must at least make sure that our process for arriving at this value is correct.
The task seems intractable. Vitaly comes up with a 70% chance of success. But what is the process behind the number? Is it just a number which he thought plausible? Or was there a process involved?
The ever quotable Mark Twain has a solution hidden inside his famous quote. “History does not repeat itself, but it does rhyme.” The answer is to look at the history of such turnarounds. As much as we would like to believe that we are unique, someone has gone through very similar experiences and has fantastic lessons for us.
One might start by answering the following questions.
How many retail turnarounds have been successful? You might try to find cases which bear similarity to the case in hand like a star CEO being hired to save an ever-fading brand. If you do not have data on retail turnarounds you can instead look for the cases were hiring a good CEO to save the business has worked (or not). This will give you a baseline estimate for the success.
What was the time it took them to become successful? This will give you an estimate on the number of years you have to hang around for the company to realize its value.
What is the track record of the CEO, if any? Always adjust for luck in this case. If the CEO has been successful once or twice, the chances are higher that he will fail this time. The luck is bound to turn at some point.
Once we have a baseline probability of success, we can adjust it according to the situation we find ourselves in. Buffett has mentioned that “retail turnarounds rarely turn” and I venture to guess that the 70% probability of success was quite optimistic. It means that hiring a very good CEO will lead to turnaround in 70% of cases.
I think a better proxy will be closer to 30%. I emphasize that I am pulling this number out from thin air and probably committing the same mistake. But this number seems more plausible because the process to reach it is less arbitrary.
Coming back to the downside, a very similar process is advisable. You start with the average duration of the turnaround and then come up with the losses/debt taken in the intervening period. This will lead to a smaller value than just taking the real estate value and subtracting the net debt.
I trust Vitaly on the calculation of the downside. I respect his opinions and have learned from his writings.
If we run the calculations for the expected return with 30% chance of success with 300% upside and 70% chances of failure with 40% downside, then the expected return reduces to 62%. If we assume that it would have taken JCP at least two years to turn itself around, a very conservative estimate in my opinion, then the compounded return reduces to 27% a year.
The investment is not so attractive anymore.
