Revisiting my American Express investment thesis.
A reflection of what I got right, what I missed, and how my process is evolving.
Last April, I developed an investment thesis for American Express but never published it publicly. I think now is a good time to revisit the decision, recognize what was right, and acknowledge the mistakes I made.
At the time, American Express was trading around $230-$250 during a period of economic stress around tariffs. What initially caught my attention was Amex’s primary revenue source: discount revenue. American Express makes money by charging a percentage of the total sale value to merchants every time a consumer swipes their card. I really liked this because, regardless of what happened during the tariff scare, people are still going to spend money and continue using their credit cards. In my mind, if prices increased, American Express could benefit because if the average transaction value increased, then Amex would make more money off every swipe, assuming transaction volume stayed the same, but I was betting volume would eventually rebound if there was an initial pullback.
One of their competitive advantages is how Amex is positioned in consumers’ minds. Their brand is built on their customer relationships and having a strong reputation for putting their customers first.
Another advantage that Amex enjoys is its closed-loop network. Building a payment network is difficult, and a closed-loop network is even harder. Even if a competitor spent massive amounts of capital trying to create their own payment network, it would still be a heavy lift.
American Express has a top-tier customer base, composed of high-income, creditworthy individuals. Until recently, they were primarily focused on older, higher-net-worth clients, but they’ve expanded into the younger generations, while still maintaining strong credit standards, to capture a higher lifetime value. This told me that the company still has a lot of room to grow without sacrificing the quality of its lending.
American Express focuses very heavily on travel, so I was surprised to see that during COVID, when effectively the whole world was shut down and nobody was traveling or vacationing, they still managed to show a net income of $3 billion. During that time, I would have expected a massive decline in total active cards, but that’s not what happened. Between 2019 and 2020, total cards in force only declined about 2% from 114 million to 112 million, which I thought was negligible considering the circumstances. To further the point, the only time in the past 20 years that American Express experienced a decline in total cards in force was in the years 2009, 2016, and 2020, with an average increase of cards outstanding of 4% annually.
Before I made my initial investment last April at an average price of roughly $243 per share, I performed a discounted cash flow analysis with these assumptions: Discount rate 6.5%, Revenue growth 3%, Expenses growth 2%, Terminal value 1%. My model gave me a net present value of roughly $242.
I thought there was a reasonably high probability of achieving an attractive return over the following year. My reasoning was based on my belief that the overall condition of the business did not deteriorate because of the tariff scare, and if consumer prices were going to increase dramatically because of tariffs, I believed American Express would be in a stronger position long term.
Reflecting on that choice, I think I was generally moving in the right direction, but I also missed some very important points:
I focused too heavily on net income, and I didn’t spend enough time analyzing free cash flow and what Buffett calls “owner’s earnings.”
I didn’t incorporate a margin of safety.
My discount rate was too aggressive. I believed Amex carried less risk than most other banks but the discount rate was still too low.
I didn’t spend enough time analyzing how management was allocating capital, and I wasn’t paying attention to return on retained earnings.
I did not clearly define what would have to go wrong for the investment thesis to fail. At the time, I mainly thought risk would involve damaging the brand or lending to lower-quality customers, but I didn’t think deeply enough about broader structural risks to the consumer credit industry itself, such as changes in regulation and potential caps on interest rates.
Looking back, I think I made a good choice with the information that I had available at the time, but I’d say I got a little lucky considering the mistakes I made and the information I overlooked. Overall, it was a good first experience investing my own money, and I’m going to carry over the lessons I learned this time into the next one.