BKNG is undervalued; AI disintermediation risk is overstated due to complexity, while share repurchases and 15% EPS growth targets support a bull thesis.

Booking Holdings is having one of its best years ever thanks mostly to AI, but also because of the war in Iran. The stock is down by over 20%. But the buyback yield is over 8.5%, and I have taken the opportunity this year to double my position in Booking Holdings.

It is today around 20% of my portfolio, my second-largest holding.

Why am I thanking AI? Because you may not have the opportunity to buy Booking Holdings at such a good price. Again, it’s rare as an investor to have such a great company at such a great price.

I believe that the market is very wrong about Booking Holdings and AI. Of course, the risk is real, but we have to be able to quantify it. Don’t just say that Booking Holdings should be down 20% in a month just because of Meta Muse.

The management has a goal. It’s called the 8-8-15 framework. They have the goal that, over the long term, they are going to grow gross bookings by 8% annually, long term. And then they are going to grow revenues by 8% annually and earnings per share by 15%.

If you see earnings per share growing by 15% annually, if they can achieve this, we can say that more or less the intrinsic value of Booking Holdings is going to grow by 15% annually.

But I believe that it can even be more than that. But 15% is a great return. Long-term returns are better than long-term average market returns.

But how possible is this? Just because management put a number out there doesn’t mean it is possible that they should be able to achieve it. We have heard numbers from some CEOs that look quite big, and they’re getting trillion dollars of valuations based on this, but of course, very unlikely that they are going to achieve this type of growth. But for Booking Holdings, it’s quite simple.

What about Booking Holdings? For them to get gross bookings to grow by 8%, we need three percentage points. Where can we get that? More people are going to be booking online. Let’s say we put one to two percentage points for online, and we put one more percentage point just for Booking Holdings, for them to get bigger market share. So you already get the 8%.

And we are being very conservative. We can get past 8%. Let’s be even more conservative and let’s say just 6%. Even with just 6% growth in revenues, I believe Booking Holdings is still undervalued.

If the company believes that gross bookings and revenues are going to grow at the same rate, it means that they are going to keep the same take rate. And this is one of the fears of Wall Street at the moment.

Wall Street believes that the take rate is going to go down because of AI.

We will talk about that in a few minutes in detail, why I don’t think it is going to happen. But the take rate has been fairly consistent long term, a little above 14%. Usually, Booking Holdings takes a 15% cut from the hotels.

This is their fee. But why 14%? Because there’s merchandising. If they are going to pay customers for them to go to the hotel, it’s contra-revenue without contribution from the hotel.

Car rentals and flights have lower take rates compared to hotels. That’s why it is 14%.

But the take rate can slightly increase because of their fintech business, where most of the payment now is going through their own platform. But more or less, I believe it is going to stay around the same.

And I think the company also believes that. That’s why they tell us that revenues and gross bookings are going to both grow at 8% long term.

How do we get to 15% now of earnings per share? It means that the margins should be improving, and I believe it can happen. But the biggest effect, why the company can grow its earnings per share bigger than revenues, is because the number of shares outstanding is going down.

The market doesn’t really understand how to value Booking Holdings. They look at earnings per share. It is important, yes, but this is not the true earnings of Booking Holdings. Free cash flow is better.

And how do I tell this? I’m not telling this out of nowhere. You just need to look at the last decade. How much money has Booking Holdings been able to return to shareholders? It is around $60 billion, mostly for buybacks, but lately also dividends.

What about the free cash flow? About the same number. So they are able to return all their free cash flow to owners.

This is not a company that requires much capital to grow, to operate. They are able to generate more cash flows compared to net income just because of the merchant business model that they have.

Initially, it was agency. You book on Booking Holdings, then they send the hotel an invoice, and they are going to be paid a commission. But now it’s merchant. They are the ones to be paid first.

So they get to keep the cash. If I pay today and I’m going to the hotel in two months, they get to keep the cash for two months. They are going to pay the hotel only a few weeks after I check out.

In the meantime, they earn interest on that, and they can use that cash for operations.

Because it is such a business with such a good balance sheet, they can also borrow money at good rates. That’s why you look at the balance sheet of the company, the equity is negative. They have repurchased so many shares.

Most of the time, when you look at negative equity, it means a company has too many liabilities, which is true in the case of Booking Holdings. But this liability, the money belonging to the merchants, it’s called merchant bookings.

In reality, it is just cash. It is generating income for the business.

It is like insurance float. Of course, we cannot compare it 100% to insurance float. Booking Holdings cannot really use that to invest in securities.

I have talked in more detail in my thesis of Booking Holdings. I posted it last week, so it doesn’t need so much update. I usually don’t post my thesis, but I wanted to be a little different.

So I posted everything on my newsletter. You can check it out. It’s completely for free, just for this once with Booking Holdings.

So let’s say Booking Holdings can repurchase 4% of shares annually. Right now, we say that the buyback yield is 8%. But let’s take 4%, which is below the long-term average, actually, because of the pandemic it even fell lower than that.

If we add that to the 8% we looked at initially, we already get 12% growth of earnings per share, but they still have 15%. But we are being very conservative.

I believe that the 15%—even if we assume 12%, don’t look at the 15%—even if we assume 12%, at the current price, Booking Holdings is undervalued, according to me.

Now let’s talk about the AI risk. This is what the market believes is a big risk. Let’s forget the war. We already talked about it. The AI risk is that Booking Holdings is going to be disintermediated.

People are going to be using AI agents to book hotels. This is the number one thing that the AI companies show, how their agents can operate.

Okay, you can book hotels because booking hotels is complicated. That’s why they want to show you that, with their AI agent, it is easy. But why is it complicated? Because there are so many players, different players involved.

It’s not just you and the hotel. First, the hotel needs a system to know if there are rooms available. They need a way for you to pay. Maybe they accept credit cards, but you don’t have one.

You live in a country where people don’t use credit cards that much, but you need a way to pay. That’s why companies like Booking Holdings exist. It’s not just for the supply and demand.

Now, can they be disintermediated? Yes, there is the possibility. But most people are still going to be using Booking Holdings because it is so fragmented, so complex.

If, let’s say, there were only 10 hotels, 10 brands of hotels, then yes, everyone can just plug into Muse or whatever, and they are going to get direct access. It’s going to be easy.

But there are hundreds of thousands of hotels worldwide, and most of them are independent. Less than 10% of the listings on Booking Holdings are actually owned by the big hotel firms.

And it’s more than just booking a room. Maybe you want a room that gets you the view of the sea. You want something better, the views of the mountain. I don’t know what you want.

The AI agent will not be able to do that. You want the reviews. You want to know that particular room, what are the reviews for that?

You get everything on Booking Holdings. You get pictures and everything. The AI agent will not be able to do that for you. It’s very, very hard. Unless you don’t really care and you just want to go somewhere for business, then maybe it can work. But not for most people.

Most people do something as travel. It is something important. You have 10 children. You need to find nine rooms. Two are twins, they sleep together. How is the AI agent going to take care of that?

Maybe Booking Holdings. They’re already incorporating AI into the business that is going to help you. But you need to have the data.

The Muse AI or the other Claude and other things, they don’t have the data unless they are going to steal it. But even if they steal it, the amount of compute they are going to use for the average person to just book hotels is so enormous that it doesn’t make economic sense for them.

Probably it’s going to get commoditized. Probably at some point they’re going to be profitable. I don’t know. It’s going to be a losing business always for the average person, and to do complicated things such as booking hotels, it’s not worth it if they actually want to be profitable,

Actually, I believe that AI can be an opportunity for Booking Holdings because they have been investing so much in AI before we heard about GPT.

I listened to all the earnings calls from 2017. The management has been talking about AI a long time ago, how to incorporate AI into the business. It is already saving them money on customer support.

But they can use AI in order to lower their cost. And they have done surveys to see that people don’t really care whether they are getting help in an emergency, whether it comes from AI or from a person, as long as it is useful.

Of course, at some point, maybe it is too complicated for the AI and you will need a person to assist you.

But let’s say you booked on Muse AI, you get a problem. Do you think Meta is going to help you? Do you think there’s someone you can call at Meta to help you on that?

As it is, I’m investing more in Booking Holdings. I believe it is undervalued. I believe that every year the intrinsic value can grow.

The intrinsic value per share will grow faster because of the share repurchases. And also, if interest rates rise, people are scared about the travel industry going down. But because of the float, I actually believe that high interest rates can be a good thing for part of the business of Booking Holdings.