Matt Peterson on His Investment Approach and Proprietary AI Research Platform

the primary purpose of this podcast is to educate and inform the views information or opinions expressed by hosts or guests are their own neither the show nor any of its content should be construed as investment advice or as a recommendation to buy or sell any particular security

Security-specific information shared on this podcast should not be relied upon as a basis for your own investment decisions. Be sure to do your own research. The podcast hosts and participants may have a position in the securities mentioned personally through subaccounts and or through separate funds and may change their holdings at any time.

Welcome everybody to a new episode of This Week in Intelligent Investing. Great to have you with us. Great to have my co-host Elliot Turner. And a special guest today, Matthew Peterson. Matt runs Peterson Investment Fund and also has...

some very interesting AI-related initiatives that he's been driving. So today what we want to talk about, Matt, is both... about your investment approach, as well as AI in equity research, the platform you're building for family offices, endowments, and other investors, which I think is fascinating.

Now, you've had a long and very successful track record in investing, and you approach it a little bit differently than most fund managers. So for those that are uninitiated, tell us about your... strategy and why it's worked so well for you over the years. Great. Thank you, John. And it's a pleasure to be here with you as well, Elliot. Thank you all for joining us.

I, after a career on Wall Street, launched Peterson Capital Management 14 years ago in 2011. And since the very inception of the fund, we've had a well... drafted process that has helped us continue to reach the pinnacle and where we're at today. We essentially run a long-term value-based strategy, and it's a very concentrated strategy. The concentration is such that we have about seven court positions that I refer to as our infinite compound or portfolio. These are holdings that we are...

confident about their growth and their potential for increased value and increased prices over time, such that we wouldn't even mind if the market just shut down for 10 years or more. These are outstanding holdings that we've...

built around and we've customized our portfolio with these holdings in mind. One of the unique aspects, I think, is that we use what I refer to as a structured value approach to purchasing our shares. And so since the inception of the fund, we have been writing cash secured puts as a tool to purchase the equity that we want to own. So we get paid up front to buy the securities and then...

We run a very concentrated long-term value investment portfolio. This year, I might mention, we're having a great year, and we're up about 60-something percent this year net. And I don't think it's a design of our portfolio to, we're not going to the office looking for 65% annualized performance. But we're sort of setting up a process where a 65% year is possible. And so we're here today and we've had teens net performance since inception.

Yes, John, it's great to be here. I'm happy to dive into any of the details. And since you mentioned such concentration, super interesting, have you talked about some of the companies that you... have in the portfolio? Yeah. In fact, we are really transparent with our investors and anybody can go and read our letters. So we have 14 years of letters, quarterly letters available on our website. And then we go into really comprehensive analysis in our annual letters. But some of them are the...

just wonderful, great companies that are compounding. And I'm happy to mention a few. We have some Chinese exposure through Alibaba. We have exposure to Tencent through NASPERS, which is in South Africa. We own great compounders in the AI and tech space like Google. And we have some long-term value-based compounders like Berkshire Hathaway.

and a smaller one daily journal also in our portfolio. And then we've, over the years, pursued a number of different advanced strategies that over time have sort of spun out and grown their own. legs and are now standing on their own so peterson for example makes a investment into the talus turkey focused fund that we launched about six or seven years ago of course the peterson

investors have access to that fund free of any fees or expenses. And recently we spun out something that we refer to as our structured dividend capture process, which is an... augmented return on our cash reserve. And we've spun that into a fund called Titan. So we also own that. And I think if I haven't mentioned them all, there's at least one more, which is Dondo Holdings, which is the Wagon Wheels Fund and other businesses that are run by Monish Pabrai, we have about a 2.3.

ownership of that business. We're not invested in the business. We are owners of that business. And so that would be our seventh of the infinite compounder portfolio that we hold at Peterson Investment Fund.

And do you always enter into your long positions via cash-secured puts, or do you sometimes outright buy when you... simply don't want to miss an opportunity yeah that's a great question because there's a lot of misconceptions i think around this topic so the first thing to point out is that we're not speculating at all we are you know

Value investors, we look at the fundamental analysis of the business. We look at the standard components from cash flow to margins to different variety of competitive advantages within the business.

And ultimately, we're just as value investors looking to pay as little as possible as we can for the underlying equity. And so I was working with Goldman on Wall Street back in. 2004 and some years around that time and basically understood just through simple algebra that it would be more effective for us to typically be selling a cash secured put going out.

any number of months, quarters, or even years as a tool to get into the equity for below market prices. So for over 20 years now, that has been fundamentally the Only way, if the products exist, it doesn't seem rational for me to just use a market or limit order like a traditional sort of independent individual investor would. it makes a lot of sense to get paid a cash premium up front as we move into the position. So I think it is in...

the 2024 annual report that I go into a lot of detail around these strategies. We talk about how at the core, the simplest strategy is to simply sell a cash secured put. But as we've... grown and expanded and practiced this method over the years, we've recognized that we get some diversification across.

strike prices and across expiration dates and so we typically build out an entire matrix of the contracts that we want to be selling we list them as limit orders in a portfolio that will sometimes sit as orders for many weeks and months before they're executed because we're looking for such extreme prices on these contracts. And ultimately... there are some holdings that do not have options available. And so when they're not available, it's...

this question of whether or not we want to own the position anyway. And then, of course, in some capacity, we do. So NASPERS doesn't have the contracts available, and we purchase that anyway. Daily Journal also doesn't have the contracts available. And our private holdings, like Dondo Holdings, of course, doesn't have an options market available. So we still do take advantage of those. It's really a matter of understanding the opportunity costs of the capital.

And whenever possible, in the fund, in my own portfolio, I talk to a lot of... individual investors about this, I highly recommend using cash secured puts as a strategy that augments your return in a value-based portfolio. So I do try to use that in our process whenever it's available and possible.

Are there any instances you could think of where you sold the puts, but the stock has run away and you never got your chance to get in? Or how would you handle that situation if not? Yes. Those situations do exist. You know, I like to think about it like this. So if we take one step back, there's kind of four high-level buckets that can occur if you're selling cash-secured put, and we can just evaluate those for a moment.

If you're writing a cash-secured point, you're basically committing to buy a security for a set price at a specific date in the future. So let's just arbitrarily say it's a 12-month contract and the stock is worth many hundreds of dollars and growing. but maybe in the market is selling for $100. So we could either buy it in the market for $100, or maybe we could sell a cash-secured put with a strike of $100 going out for a year.

request a payment of, say, $20 for that contract. So what that would mean is that we're holding $80 of our cash when we get paid 20 more. And then we hold on to that cash for the duration of the year. And at the end of the year, there's a binomial outcome based on the stock's price. If the stock is above the current strike price on our contract,

It's out of the money, and the contract's over, and we simply keep the cash premium that we were paid. And in this case, in this example, it would be $20 on an $80 collateral, so it would be a 25% return. If, on the other hand, the stock is below the strike price at expiration, the stock would be put to our portfolio and our net cost would be $80. So whatever the stock is trading for at expiration of the contract, we would be buying for $80.

rather than the $100 when we felt like we wanted to move into the stock. So as I mentioned, there's basically four categories. The first one is the... Worst of all, which is that we are totally wrong and the stock goes... Ultimately, it could go completely to zero, but more realistically, it just declines or falls in price. And so that's a very bad situation. But let's take an example. If the stock were to fall.

50% after our analysis was complete, it would of course be a terrible situation. But instead of spending... $100 on the stock and now having it at $50 and determining what to do next. We would have paid only $80 for the stock and the price would be $50. So instead of losing $50, we would lose... $30. And so that is a superior outcome to just buying the stock with a market or limit order in the market.

The other two buckets are greatly preferred. So one being where it sells just trades for just slightly below the $100 strike price. A year from now, we've picked up $20. And so we're buying for $80 and maybe the stock selling for $90 or $90.

we've already made a decent return and we have the stock in our portfolio and our net cost to purchase the shares was only $80. And I think it's worth maybe in this scenario pointing out that if we're looking at a long-term compounder and we're analyzing it let's say for maybe some easy math that we're hoping or anticipating that in a series of years the shares would be trading for eight hundred dollars per share maybe the value

is today $300, but it's growing. And we expect that we might have an exit around $800 a share. Well, if we buy for $100, we'll have basically made 7x in our portfolio and on a return. And if we buy for... $80 and sell it for $800, we'll have made a 9x. And if that happens, whether it takes 10 or 20 years for that to occur, that difference in the IRR and the performance is substantial.

And I will also mention, even though we're talking about these buckets and this bucket too, when we hold the shares, We also exit with our structured value practices. So this is like value investing combined with structured products like options. And when we would exit... the shares, if we ever have to exit, we would exit selling a covered call against our shares. So even if it were to appreciate to $800 a share, we would then take that opportunity to write a covered call against the stock.

pick up another 50 or so dollars and exit above $800 a share. So we get a better price going in and a better price going out. And if on average, each of those were to provide, say, 2% annualized. we end up with a 4% annualized boost because we're using the structured products in combination with our value practices. So that kind of explains like... The one worst scenario where it goes all the way down. And then I think the best scenario where we actually.

obtain the shares and they've sort of lagged the market or they've stayed down a little bit. The third scenario, the third bucket is when the shares are trading slightly above the strike price. So maybe the shares go to 105. or 110, and we've picked up $20 for the contract on our $80 in collateral, and the contract expires, and we don't pick up the shares. However, because the shares haven't moved much in price, we're in a position where that's still the optimal...

And maybe it even looks better because there's been more time and the business has developed over that period of time. Maybe we have additional confidence. We can go ahead and sell additional put contracts once the contract expires.

do the entire practice again as you mentioned elliot the second to worst scenario is that a company basically goes up a lot in price our analysis is correct and the stock completely runs away before we are able to get the securities and then we maybe collect the 25 percent gain in this example. We still keep the $20 in premium on our $80 in collateral, but we've now missed the shares. And if the shares were to go from, say,

$100 to $300, of course, we would have had a better situation if we had just bought. Now, it's interesting because this has happened on a few occasions. And we do tend to have really outstanding companies within the portfolio. However, it doesn't happen. As often as you think, I think in terms of value investors in general, we're also patient and looking for these unique opportunities. A lot of times what we find to be unique are.

are companies that have sort of off financial statement value, or there's a narrative that the public hasn't really picked up on yet. And sometimes it takes longer than we anticipate for that shift to happen. We sort of look at a company that's very mispriced and question how this could be.

And so we have had, you know, in some scenarios where there's some operating leverage, maybe there's a little bit of debt. We've had some companies that go up and maybe we missed the opportunity, but there are. many opportunities. There's 6,500 companies that we could buy in the U.S. alone. And so if a company kind of runs away, it's a little bit disappointing that we miss something.

But it's really not the end of the world. We've collected a 25% yield over the course of the year. And there's usually a new opportunity that's available to us. So we'll put the capital back to work. in something that's really timely at that moment. So we haven't had any... complaints about the things that we've missed because we tend to get the premiums and the shares are going up and the portfolios on the rise. So the most notable

And what we're always concerned about is our downside risk and our downside exposure. We never want to be in a situation where we're in bucket number one and the shares decline. We try very hard to avoid. owning melting ice cubes or things that will sort of devalue over time. We want time to be our friend. And so although we do are aware that things can run on us, things tend to be a little bit volatile.

And we are usually pretty early in our analysis and understanding of these firms. And so it hasn't been a major issue. And there's always a new opportunity around the corner. And you've been doing it long enough, Matt, that you've seen all kinds of market drawdowns as well. So I think that's testament to your risk management.

process there as well. And certainly COVID must have been an interesting time, but I think you did quite well in 2020. Yes, thank you for that, John. That is correct. We've had this in practice since the very beginning.

One thing that I will note again, back to this 2024 annual report, I go through six different diagrams where I've sort of shared the evolution of our thinking in this process. So at the beginning, it was... Pretty plain vanilla. We were still sort of using the matrix approach, which has all sorts of psychological and behavioral advantages. But, you know, we were just kind of selling puts at the core.

and then waiting for the stock to be put to us so we could build out our infinite compounder portfolio. Over the years, what we found is that there are a number of scenarios that give us some opportunities in addition to... to sort of the plain vanilla strategy that I do recommend everybody start with. And I think one of the...

biggest hurdles is for somebody to just go and try it. You watch the, you know, if you go into your individual portfolio and you set up your permissions so that you're even allowed to. take on this type of exposure, this cash-secured put exposure, you're half of the way there. And once you write one, recognizing that your notional exposure is 100 times the number of contracts that you've sold.

you'll basically watch this unfold as it moves to expiration and it will become very clear. So for anybody in the audience who... you know, doesn't think they understand this, I really recommend they look through their fundamentals. They find the great value opportunity. And instead of buying 100 shares with a market or limit order, they just go out and write one put.

Now, some of the things that we've discovered get, I think, a little bit into some complexity, but it's helped us, you know, especially in this year where we've had this 65% run. What we found is that sometimes the price doesn't match the value in the market. And I think that's probably common knowledge for a lot of your viewers. But what's surprising is that these...

shares that are already cheap could get cheaper. And so we've sold puts at times. I'll go into a much more advanced strategy. So one in the middle that some of your viewers might. understand is sort of called the wheel and or it's referred to as the wheel and what that really means is that you're selling puts on a stock that you want to own, and then later after you've owned it, you go and you sell calls. And that's the wheel. That's cash.

Through the cash-secured put, you end up with stock. And then through the sale of a covered call, you end up back to cash. And so that wheel is an effective way to basically move in and out of securities. realized over time is that first of all there's a systematic mispricing of these contracts so there's just much more demand than there is supply so you get a higher price than is

really rational. So we're happy to be sort of the insurer because we know the qualitative aspects. We know the fundamentals very well. There's somebody on the other side of the contract that's looking for some protection, and we're willing to provide that to them for a fee.

And then, of course, if the shares decline, we end up with this stock. What we found is that the shares can and will decline at times. And if the thesis still holds...

it presents an even greater opportunity. And so we employ some sophisticated tax harvesting techniques where before a contract will become, say, a long-term... capital gain or loss there's this period where it's a short-term capital gain or loss and so for any u.s investors they understand that it's it's valuable to be able to write off the short-term losses before they flip to a long-term loss because it's written off as

income at your marginal tax rate. And so one thing that we've found to be quite attractive is that if there is a decline in price... we can actually unwind the short puts before they flip to a long-term capital gain. And in those cases where the thesis still holds or even looks more promising,

We've taken a small amount of capital and bought some long-dated, out-of-the-money leap calls. And that gives us a... little bit of an opportunity if the shit, and then we will resell new cash secured puts and we will bring in a bunch of premium and we will use a little bit of that premium to buy some deep out of the money.

long-dated calls. And what that does is it protects us from a company really running away. And that's actually what happened. I think a lot of people followed Baba. They were watching maybe when Charlie Munger was purchasing it. He was buying it inside of Daily Journal. I think around $200 a share was his entry point. Our original puts were being sold with strikes of about $100 a share, and we were picking up premiums on that. And we found that as those were expiring...

The price was in the 80s and in the 90s, and at some times it was even in the 70s. And so that was a situation where it was quite clear to us that there was a major mispricing. And we could have just simply sold cash secured puts. But this year, quite, I think, suddenly over the last 10 months, the shares rallied from about 80 or so. And I think peaked at 190, and maybe today it's in the 150s or so. So that's a situation, Elliot, where we would have missed Alibaba. But because we had...

put this more advanced approach where we were buying some out of the money calls at maximum pessimism when it looked really unfavorable and they were extremely inexpensive. We were still able to capture the shares on the way up. We paid for them with our cash-secured puts that we were selling on the way down. We have not realized any capital gains on the transaction, but we have taken...

short term tax write offs for our clients. So we've just picked up like two, three, maybe four different really positive aspects. with this one company and transaction. And all of this has taken about, I think, three to four years to unfold. And you took the short-term losses on the puts?

Correct. So what happens is if you sell a put, let's use the old example since the price is... are similar. So if the stock, you could sell a strike price in, out, or at the money. So at the money would be that it's at the same price as the stock price today. So let's take an at-the-money example where the stock's trading for $100 and we're selling put contracts for $100 and we pick up $20 in premium. If the stock price falls to $80, the general market

And the general market is not valuing every company that they come across. Some are trying to do it. Others are traders, speculators. Others have a more complex portfolio construction. And they are using these. Maybe they're trying to fill a bucket in the IT or international space. And as those shares drop to 80, the put prices go up. And we are essentially short.

that put we've sold the put so if we sell a put for 20 while the stock's trading at 100 and it falls to 80 the shares might the puts might go to 30 or 35 dollars And so what we're able to do is unwind those at a short-term loss of $15 per contract. And then write a new contract with a different expiration date. So basically moving the put to another expiration date further out into the future.

And so we capture the tax write-offs. Those go to the K-1s on our LP's tax filings. And then we end up selling new.

put contracts, picking up new premiums, and then at the same time using a little bit of that premium to buy the out-of-the-money calls just in case the stock does run away. And given that it was at sort of the maximum... pessimism the chance for it to say move by 30 40 50 percent becomes relatively reasonable i mean this is like a almost normal fluctuation for an equity a single equity in the market can swing by 30% and 40% a year. So when we get that opportunity where...

We've seen some declines in the stock. We've already sold the put. We can capture those tax write-offs, move the contracts to a different expiration date, grab a long-term call, and then ultimately end up as owners of those shares. became become long-term unrealized compounding engines for the portfolio and just to understand it so is there no like wash sale rule for options or is it not there if you're doing a new contract with a different expiration?

Yes. So they are, it's not there under a new contract with a different expiration. So there are so many different option contracts. There's so many different ways that you can make this happen. And the wash rules are really. based more on the Q-sips for the equity holders. If you, I think, were to buy and be trading...

an individual option contract. The wash rules, I believe, should apply. But these are sort of like different contracts, different expirations. And because we're also long-term thinkers, Many things are not always happening within the 30. days that would violate the wash rule. So we may be building into a position over a six-month period, selling contracts periodically. We may be unwinding them also.

over a period of time and then buying into new ones over a period of time. So we are able to capture those write-offs when we unwind a... a put that has increased in value that we've sold and that has increased in value. We're able to unwind that and capture the short-term write-off and then just move to a new contract on the options chain.

And in terms of the discipline required to execute this strategy, I would imagine it requires a ton of discipline because you've seen such success. I mean, I know I would be... Maybe a little bit tempted to not just do cash secured, but to say, okay, we can lever this up a tiny bit, you know? Sure. Is that something you had to just struggle with and then decided?

We're just going to do it this one way. Does it ever get to where you feel like you could do something with that cash that's waiting there? You know, it's a good question. I think there's a number of ways to sort of... analyze the situation. So it does require, I think, a superior amount of discipline. However, I don't think it's discipline that other value investors lack. It's just a mindset where I truly look at ourselves as owners of the shares.

And because we own the position, I just sort of, I look into the future, say, 3, 6, 9, 12 months, and I think this cash has already been... you know, this cash has a purpose. And it's to say, own the underlier that we've sold puts on. So there is a little bit of temptation, I think, but it's... So, you know, there are things that we do to sort of augment the cash. So for many years, it was a real challenge when interest rates were zero.

Because there's cash sitting there that's sort of getting eroded by inflation. And even though it's going to be put to use somewhere, maybe there's something that can be done that's a better use for that cash. And so maybe there's a special situation that would certainly occur like a merger arbitrage sort of opportunity where we could do a little something.

But ultimately, we just need to have the cash available. And it doesn't end up being a lot of cash in terms of the overall portfolio structure because we already have... accounted for the other holdings in the portfolio. Those have already converted to equity. So it's the cash affiliated with like one single holding. One thing that we started to do over time.

that really augmented the returns was something we called structured dividend capture. And I've talked about that extensively in some letters and even on a few other podcasts. But ultimately, what we used is some... AI tools along with some components of our process that we've automated to identify this sort of superset of opportunities that you could sort of think.

are like on our watch list and that oftentimes will pay a high dividend and that we have options affiliated with the underlying equity. And what we've found is... It's quite valuable if we have a very long-term put contract that we're waiting for. We actually can sort of run a wheel. which is very short term, around the ex-dividend date of a very stable company, and make really great returns. So sometimes that cash isn't sitting totally idle.

We're doing a little something to augment the returns. We just kind of want to outperform sort of the interest that we'd be receiving on that cash anyway. But we found that we do get quite a bit of return from that.

actually spun that out into its own entity now called Titan. And so Titan is pursuing this. So now when we have cash, we actually will sweep it into Titan. It will be managed conservatively at Titan. And then when we need the cash back to purchase the underlier, we just sweep it back into the Peterson account, and then we use that cash to buy the stock. And so in my mind, the discipline is...

It's very similar to a standard equity portfolio, value-based equity portfolio with long-term concentrated holdings. And it just becomes a discipline of not over-trading.

the sort of contracts, you know, you can buy these, you can sell these. It's not like you just sell a put and then you have to leave it alone. So the discipline comes with not making many adjustments to sort of going back to your thesis making sure it still holds and leaving the portfolio to sort of bake to mature and ultimately ideally you know you receive the equity that you set up to obtain

So Titan, it sounds super interesting. That's the liquid cash management fund you launched in October, I believe. That's right. Does that have basically daily liquidity or how does that work that you can sweep it right back in when you need the cash? Well, so it actually has quarterly liquidity. So we have quarterly commitment dates, quarterly redemption dates. However, there's a few things I think to understand. So first of all, the contracts that we use short secured.

cash-secured put contracts, these have a defined expiration date. So we can see far in advance when we're going to need cash. We also custody our assets with interactive brokers that has i think the best most competitive rates both in the interest that they pay on cash and the cost of any margin exposure and because basically our portfolio is already full of these infinite compounders. Our portfolio is about 40 million AUM, not enormous, but we started at them from a much, much smaller base.

Actually, just $100,000 from myself and $25,000 from a friend 14 years ago. So now we're at 40. If we have exposure to, say, a position that's going to take a 10%... holding, that's really only $4 billion. And so if there's a slight mismatch of when we get the cash versus when we buy it, I'm not...

opposed to having a small amount on margin. I don't want to have extended periods of margin. We're not trying to own things on margin. But if there's, you know, a contract that expires January 19th... and then we know we're going to get our millions in from Titan April 1st, I'm sort of okay with waiting for that two months plus to...

balance all the books and then Titan's performance will offset the interest that we pay on our margin anyway. So it doesn't need to happen on exactly the same day. Titan has quarterly liquidity. And we find that sufficient for our needs. And we have external investors that are participating now in Titan, even though it's a brand new fund. We've opened that up.

externally as well and so i think quarterly is manageable for everybody and it basically we we spun it out because of demand i mean because people kept asking me, how can they get exposure to what we're doing with our cash without getting exposure to Daily Journal and Berkshire and Dondo Holdings? And ultimately, the only way to do that was to separate the entities. And so we spun that out, launched it independently. And we have a portfolio manager who is solely in charge of those assets.

It does a number of things that we might want to get into, but it frees up my mind to focus on the research and the long-term compounders without the distraction of... thinking through a few weeks of sort of some faster trades. And so we've set that all up. It's available both for our cash and for external capital.

So let's shift gears and talk about the Veritas Alpha, the AI equity research platform you've put together. Why did you... do that what does it do what's you know what's the purpose of it and how is it different from someone using gemini let's say okay well i'm very excited about what's happened with veritas alpha

Ultimately, what happened? Why did we start this? And what is it? Okay, these are all great questions. There's a lot here. So what this is, is basically an analyst. that delivers real-time analyst reports to my inbox. And we can run this now on any company that files with the SEC in America, and we can even run it on any company. India, and we can turn on any country where there's adequate data. And ultimately, all that's required is you put in a ticker, put in your email address.

20 minutes later, it runs through thousands of lines of Python code. It goes out to the various LLMs, and it builds. An analyst report that is of the highest quality, institutional quality, and it's 20 minutes. They run in parallel. I can order entire sectors. order entire countries so it is absolutely remarkable i feel like it's sort of the answer to a lot of us about our value investor prayers and the reason is that the reason i set it up the reason i built this

was quite literally for myself. It was, you know, back in 2020 during COVID period, I started trying to automate using Python different. functions that were repeatable. It doesn't replace judgment.

It still requires human analysis. But what it does is it does all these shortcuts that combine all the information, maybe produces the frameworks that I always want to see, and it truly acts like a... junior analyst that i can then assess and and evaluate so you know us value investors are are always just going through you're on this information treadmill

The world's always changing, and once sort of critical mass learns something, it's no longer of any value, and you've got to move up the chain and learn something new to sort of add any alpha to your portfolio.

And it never stops. It's just like, you know, you have to keep up with, and it's quite literally, I think, impossible. I mean... I don't know that many fund managers would admit it, but how could you possibly, with a handful of analysts, maybe some great friends that are portfolio managers, John, you've been all over these companies for so many years, you've run all these conferences, and still there are opportunities that are missed by the whole community. How do you keep up with...

just the 6,500 or so companies in America alone. A single analyst on Wall Street might be in charge of a couple of dozen. Just keeping up with the earnings calls and the changes, it never ends. And so... You know, for my own interest and the interest of having our great returns for our partners, I wanted to automate everything that I could. And I had learned that Python was going to be the language we should use to do that.

We started by almost recreating whale wisdom. And I know we've talked about this over years, John, but internally, it was like, let's figure out the superset of what all of our favorite long-term concentrated value investors are holding. How many companies are they holding? And so we have this process where we evaluate about 100 different funds. And interestingly, we rarely see that they...

collectively own over about 500 companies within the market. So we almost immediately can eliminate like 90% of the US market. If it's not within those 13 Fs or there aren't crumbs leading to some other entity from a pool that's already owned by... these great investors, you know, maybe there's something that's worth a little bit of our time, but, but possibly not much because we're looking for needles in a haystack. It's not, you know, most of the time.

The market is pretty efficient and the prices are fairly accurate. And you need to find something that's obscure. There's off financial statement value or there's a very distinct reason. why something's mispriced or really out of favor, whether it's the entire sector or individual company, there needs to be a reason why it's wrong.

Those needles appear much more frequently in the pool of assets that are owned by all these great value investors. So we're looking at Bill Ackman and we're looking at... Warren Buffett and we're looking at David Einhorn. We're looking at all the names that you would think that we're looking at. We're looking at what they're buying and what they're selling and how many of them own the companies and what percent of the assets owned by the great investors.

You know, if we go off and outside of that concentrated space, that superset, we find fewer needles. And it takes just as much time. So our time is much more effective inside of that space where there are other value investors that have picked something apart. found something and maybe with a multi-billion dollar portfolio allocated you know 10 they've put

thousands of hours into their research. They have huge analyst teams, etc. So it gives us a real shortcut. I think it's a real advantage for us. So we started automating those types of things. And we were doing well with all of that. And we are putting in place some of these structured dividend captures. We are trying to grab dividends from that superset of companies using a wheel, cash-steered puts, and covered calls around the dividend and finding some great...

you know, results with that. And then there was the whole chat GPT moment that I think everyone experienced at some point I experienced on the first day. And it was like, I mean, it was truly like. like the answer to my prayers. It was like I had just been given an infinite number of analysts that don't need a vacation and don't need to sleep. The challenge is it's tough to keep them all going without some...

technology and code in the background. And so what we did is we started building out this really robust platform. where it would run analyst reports. And I encourage all of your viewers to go there. It's all free right now, which is really remarkable. I mean, we are just doing it. all for free. It's publicly available, and it's on a website called VeritasAlpha.com, and you go there.

There's not even a login at this point. You just put in the ticker, you put in your email address, and 20 minutes later, you're going to get your analyst report on that company. And it goes through as much as we can reasonably include. and find to be very accurate. We have all sorts of frameworks like SWOT analysis, Porter's Five Forces. We go through an entire moats framework. We do a sector analysis.

we go into extreme detail and it required us to get like API access to the SEC so that when you type in the ticker, it can figure out the QCIP, it can download.

all of the past annual reports, quarterly reports, any information of relevance, and basically take the action that an analyst... would traditionally take so to write one of these reports it consumes thousands of pages it goes through and it there's a reference page at the end where it tells you everything that it's looked at it goes through and says well we've read these six ten

Ks and we've read these 2010 Qs and we've gone through all of these earnings calls and here's your report and it starts comparing competitors and it does valuation analysis. bull bear base case and even writes up an unbiased short thesis. So I think that writing one of these high quality reports for any level of analyst requires weeks of work.

just consuming the information alone and coming up with the possible scenario analysis. And in 20 minutes, basically I can, in 20 minutes, I can have the results.

And there are about 50 or 60 pages packed with great information. I can go through that report in 10 or 20 minutes and look at it as a cue to whether or not it deserves a further look. And so now I'm... constantly the breadth of what we can explore is so much more vast and the depth that we can get into quickly is very very high and so this has been like a

Just incredible opportunity and advantage for us. So we're now working on making it even better. We've always been improving it. We're like trying to make these reports themselves compound. We've had quite a few users a year ago. Nobody knew that we built this, even just, I think, six months ago. Last year at the Berkshire meeting was the first time we really put it on an open website. Before that, we were just using it internally. And we started with a few orders per day.

Now we consistently have hundreds of orders per day. I think the max number of reports ever ordered was 1,800 in a day. We had another day with 1,400, so we've had many thousand-plus order days. And ultimately, we'll put it into a SaaS business and have a subscription license for it. I don't think it will be cheap. It will be probably like $9,000 or $10,000 a year.

But right now it's all free and it's going to remain free for at least another quarter or two because there are still some bugs and things we are working out. There's not many hallucinations anymore, but there are... Times when maybe it doesn't grab the very latest quarter. And we just don't want to deal with paying users until we work out all these bugs.

But it is extremely valuable for our own research capability. It allows us to just simply turn over way more stones. We get to look much further than we ever could before. The analysis is very high quality. And it's really, really something I use on a daily basis. Yeah, go ahead, Elliot. Yeah, I'm curious, kind of two dimensions of questions. One on...

how everything works and the other maybe changes in your own investment philosophy that might be born of this. But have you basically built like a series of coded prompts? that are designed to ping edgar and it just focuses on edgar and the actual reported financials of the company to pull the report together is that the kind of source of truth

that you're building on and it's about the prompts and how you've engineered them. And the second question, it's kind of related, but you talked about being able to cover way more companies and more quickly say, you know, whether it's worth your time or not. Does that maybe open up the possibility to broadening the number of positions in your portfolio and being, instead of concentrated, maybe semi-concentrated if you could cover that much more range?

Okay, great. Both great questions. Let's take the second one first because I think the answer is a bit faster. I still look at the portfolio in terms of opportunity costs, and we're still looking to maximize the long-term performance. of, of the fund. You know, I think we're up about 500% since inception and, and I'd like to see a point when we've like done a hundred X on capital. So, you know, it's like we're five.

percent of the way there we have 95 percent of the way to go i look at this very much in terms of opportunity costs and i do find yeah we can we we will be able to turn over more stones But then we can also identify which ones are better opportunities in terms of risk-reward. So I don't anticipate us having ever a much broader... I sort of look at things from a Kelly criterion mathematical perspective, and I find it easier to monitor, you know, maybe 10 or 12 positions, not 25 or 30.

And I typically am of the mindset that if I'm looking at the 30th and I can't determine whether or not that's better than our first, I think that's a problem. in itself so i'm not necessarily looking to expand broaden the number of positions in the portfolio although over time we will introduce more and we will

not necessarily liquidate existing infinite compounders if they're still growing and the thesis exists. So I do expect that we will get more a little bit over time, but I don't ever think we're going to have a, you know... super diversified portfolio or like a mutual fund or we're not like closet indexers at all. We are very, very unique and different.

I think the short answer is we can turn over more stones. We look at it like the opportunity costs. We evaluate them against each other. And we still want to have only the very best, you know, the best management teams in the world. We want to have the best business models. in the world we want to have buy in at the best price as we possibly can and so that that philosophy still exists we just get to look at more and so if i go to a conference

and there are, say, 60 or 70 companies that are presented over a few days by a bunch of really brilliant people, I can simply run an analyst report while they're speaking. And at the end of the day, I have 20 or 30. in my inbox i can flip through them i can get to know the companies very well i can look at the bull bear base short thesis analysis i can look at all the scenarios i can evaluate the risks look at the competitive advantages and quickly understand like

Is this a company that should be in our top priority of future research? Or is it a company that we've now glanced at and we can... put on the back burner. And so it's more that we can turn over more stones. We can see more. And then because of that, because we're just looking at more opportunities. As long as our analysis is correct, we're going to find better opportunities. And so that fundamentally is what we use the platform is to really broaden.

and deepen our research very quickly. It's about speed. And because we're able to do that, we're able to find the best opportunities for our portfolio. Now, how it actually works is it's...

much more advanced than what you just specified. But at a high level, that's what we're attempting to do. We have Different resources that we classify as Tier 1, Tier 2, and ultimately Tier 3. And the Tier 1 resources are the factual resources, like you've mentioned just from Edgar. So we go to Edgar, a few databases that are, you know, just have financial data available. And we pull up all of this information.

But we've also built these foundational documents. I've built these documents over years. And it's things like, with the help of AI, I've written like 30 to 50 page. analyses of a sector and so you know you would treat company in the energy sector much different than you'd treat a REIT for example. And so we have metadata on the back end affiliated with every company that somebody would request from the system.

And the metadata isn't hard to obtain. This is like factual things, the market cap, the sector that it's in. And then I have underlying documents that say if it's in this sector. Don't forget to go off on this branch. and do this type of analysis. We need to get to unit level economics in some cases, or we need to look at the core commodity inputs in other cases. There's different things that happen. And so it's not as simple as just going to like Gemini and saying,

Tell me about Amazon. That's not going to give you a very good answer. It's going to give you something. It's not going to really give you an investment thesis. It's not going to compare across the competitors naturally. You're not going to be able to trust very well a...

sort of dcf analysis but what we've done is we've created these foundational documents that are in our system that everything sits on top of and so we start with sort of this this core It's like a tree with branches, and we start with the trunk, and we build out everything we can around. The perfect prompts, you know, the prompt engineering is one aspect, and it's a very important aspect. And what it is, is it's sort of...

If you run the reports and you do a lot of them, you'll start to see that it's kind of component-based. So we need to understand all of the different frameworks. We need to understand the full risk analysis. We need to know which... components are going to feed into this sector, this market cap level of a DCF, or should it be a NAV evaluation? Should we be looking at

a variety of different cash flows. How should we treat this company? And we need to do it in a way that's generic enough that it can scale. And so we've engineered the prompts accordingly and we will spend an enormous amount of time like... just getting a SWOT analysis to come out properly to evaluate, you know, not only the obvious.

situations that you know any analyst would go and pick up with a google search but also trying to identify like the less conventional scenarios and then evaluate whether that is really a truly realistic possibility or if it's just a noise. And so we've taken a lot of steps to sort of incorporate these things. I think one One interesting thing that we've done, and we're getting even more advanced than this now, but one interesting thing is that every quarter...

we round robin the LLM that we're using. And we found that that really helps to eliminate the hallucinations. So, you know, at some point we were running it all on ChatGPT. We've moved everything over to Gemini. At some point, we might be running on Grok. You know, in January, we're either going to be on Anthropic, Claude, or we'll be running on Grok. And then ultimately, what we will get to is a sort of decentralized... where we're going to use a framework where we just have an agent.

that identifies the best LLMs for that particular component of the report. And so it will be actually a report that's produced by multiple LLMs simultaneously. Everything gets wrapped together. in a really nice, I guess, sort of, you know, template, let's say. And it comes out looking currently like a very professional, like Wall Street level. analyst report that you'd get from a junior analyst or maybe even a team of junior analysts or a

more senior analysts. It's really quite profound. And so when you go through there, maybe you'll check a few numbers at first. You'll start to realize they're all correct. We go in great... depths into the earnings call we try to pull out components that would analyze like the quality of this management the different decisions they've made over the past we really get into a lot of depth and so we'll spend as i mentioned like

a series of months, fine-tuning a specific framework and then incorporating that into the report and scaling it across. the entire country. And everything just gets better. And every time the LLMs improve, the analysis gets better. I think within a couple of years, we're going to have an AGI analyst. And I think it's not going to replace, it's not designed to replace.

analysts or portfolio managers that's not the intention it doesn't replace judgment it just allows you to turn over so many more stones get deeper insights, evaluate things like, you know, you can just incorporate your checklist into the analyst report. And now you don't have to spend a day going through the checklist. The checklist is done and it tells you what's green and what's red.

you can just move forward to the next company you're looking to evaluate. Terrific. And Matt, the website is veritasalpha.com, correct? The website's veritasalpha.com. You know, right now it's sort of a beta version. You can also get there. I have a beta version on the petersonfunds.com website as well, where you can see letters and things. But Veritas Alpha is where...

You actually just go in there. You hit a big blue button. It's so simple. It's almost shocking. You type in a ticker and your email address, and you will get the report. Terrific. Well, this has been a fascinating conversation. Thank you so much, Matt. I really enjoyed it, and I hope our listeners got a lot out of it as well. Thank you, Elliot, also for your questions.

And thanks, everybody, for listening. Matt, truly a pleasure. Thank you. Thank you, John. And thank you, Elliot. Thank you all. It's a pleasure to be back. And yeah, I highly recommend a few of these, taking advantage of a few of the topics we've discussed. Really go out and write your own set cash-secured puts. We explain a lot of it in our letters. And you can, I highly recommend, especially while it's free, going to VeritasAlpha.com, typing in some tickers from your own portfolio.

and looking at the results. Great. Thanks, everybody. Talk soon. Bye for now. Thank you for listening to This Week in Intelligent Investing, brought to you exclusively by MOI Global, the research-driven membership organization. Learn more at moiglobal.com.