Man vs. Machine Cat Bonds High Yields, High Risks, But Are They Worth It

In our latest installment of Man Versus Machine, we asked Chat GPTA personal finance question, what are catastrophe bonds and what are the pros and cons associated with those bonds? And then we asked the subject matter expert to critique that answer. What did the ChatGPT get right? What did it get wrong? And what were the material omissions? And our subject matter expert for this edition of Man Versus Machine is Brian Regan from Wealth Enhancement Group. Brian, welcome. Thanks for having me Bob, as usual, it's always a pleasure. Thanks for being here. Thanks for volunteering to critique chat GP TS answer about the pros and cons of catastrophe bonds. Where are you going to have you walk through what you believe it got right, what it got wrong and what were the material emissions? You know, I always guess I'll just jump right into it. ChatGPT did a fine job. I would say giving a giving an overview of you know, high level. If you knew a little bit about what they were, you know, you'd be able to follow along pretty well. For those of you who are unfamiliar, Catastropheree bonds are typically issued by insurance companies to cover some of their property and casualty liabilities. So basically it puts the risk upon the investors who are interested in in buying these of of catastrophe these happening and and causing a a large payout to the insurance company. So, you know, ChatGPT gave us a list of pros and cons and, and we'll start with the pros. So it said that they have high yields. That's true. They they typically have yields, You know, north of right now, it looks like north of nine percent, they have maybe maybe 3% higher than an unsecured high yield bond right now. So you know, you're looking at 10:10 to 11% right now, which is, you know, obviously pretty attractive, right. So there's there's no doubt about that. They do have higher yields diversification. I think this is also true. You know, they can't provide diversification because if you have a conventional stock bond portfolio, catastrophe bonds aren't going to trade like your regular stocks, either investment grade or high yield. They're really going to be contingent upon a trigger. So another trigger event can be a massive hurricane or some kind of large claim that comes on because the insurance company, you know, experiences some pain limited duration. So this is one of the places where I start to critique ChatGPT says it's a pro. Technically it's correct. It's, you know, duration around three to five years when they're, when they're issued. But the duration isn't really a, the risk for, for this type of asset, right? The, the, the, the risk is whether or not we're going to have some kind of massive, massive catastrophe that's going to wipe out the principle of your balance. And, you know, the assessing that risk rather than doing, you know, a yield curve analysis is, is really what's going to be important here. So, you know, I didn't really find that one to be relevant for this type of investment. And the 4th one is, is trigger events. So the payouts linked to a specific event. I found this interesting as a pro, you know, this is, I guess this isn't a pro for anybody. So let's say some massive thing happens and there's obviously victims that, you know, come from that. So it's not, it's not good for those folks. Now, let's say they're insured. The, the insurance company is kind of going to going to be made whole by the, by the principle of the fund. So it's a, it's neutral for them. It's not even a pro for the insurance company. It's a, it's neutral and it's certainly a con for the investor, right? If we do have this kind of massive event, that's how you're going to lose money on this, right? That's why you're getting the higher yield. Now, I, I did write in the article, you know, the high yields could arguably be, you know, a con as well. It's, it's, it's certainly a, you know, you're getting those, you're getting those high yields for a reason. The reason why you're getting those high yields is because this is a very difficult thing to assess. You're going to need a team of actuaries to determine the probability of this event and pricing that correctly could be, you know, a significant challenge. In addition to that, you're going to be at the whims of, you know, some kind of third party facilitator of determining the amount of claims. And that could be, that could be squishy. You know, there's, there's squishy events and, and formulas that that play into that. I've read about many cases that, you know, 11 certain variable was flipped from, you know, a pro to a con and it wiped out the whole fund. So that, that's something to, to, to consider. Now the cons are high risk. You know, we, we talked about some of that, you know, just the fact that we could have a catastrophe and you lose all of your principal. It's very rare when you make an investment in your likelihood of losing all of your principles is high. And again, that's why you have, you know, the high yields. Now it's, it's important to note, even you would have to have a very significant event to lose 100% of the principal, right? Because you're only going to take a hit on the amount of the claim. So let's say there's a, you know, $100 par bond and there's $10 worth of claims. Well, you're still going to get $90.00 back. So that could be a good deal still, right? If you're getting 10% a year over four years, you're going to get $40 back, you're going to lose $10 in principal, you're still $30 up. That's not an entirely bad deal, but there, there are cases where you could lose $100.00 of your principal back. And even if you're getting $10.00 / 4 years, which the likelihood of that is, would only happen if the, the, the trigger event happened at the end needs to still only get, you know, 40-40 cents on the dollar, which you know, is a, is a tough, is a tough thing to swallow complexity. We talked about that there's going to be, you're going to need some actuaries to price this risk and even, you know, ask any insurance company in the world. Actuaries aren't, you know, that's it's a statistical model where you hope for the best and the world's ever changing. So the those statistical models are based on backward looking evidence. When you know that the things moving forward seem to change pretty fairly regularly, particularly when it comes to catch traffic events. Liquidity liquidity is a real problem for these you if you enter into this the the expectation should probably be that you're going to hold it ChatGPT got that right What chat P GPT did not get first of all, didn't say anything about actuaries right. So we could start there with liquidity. There are certain times when liquidity is better. So you know, it's not necessarily good during hurricane season, but maybe right after hurricane season it it might be, might be better. So there's certain times in the calendar year or certain times in the catastrophic event year that it might be better. When it comes to complexity, I'm not quite sure if we mentioned this, but you know, the issuer is going to lay out the event and the parameters of that and they can be as simple or as complex as they want to make it. And if the market is high in demand for these type of bonds or starving for yield, you know, they they might be a very issuer friendly type bond. Just just things to consider. You know, it throws in regulatory and tax considerations, which every investment has that. So it's it almost seems like a kind of a throw away to consider. But the biggest thing, and I think it makes mixes up on these pros and cons is is it a pro for the investor or is it a pro for the insurance company? Is it a con for the investors? A con for the insurance company, right. When it comes to complexity, could be a pro or a con depending on what party you're representing. I know when it comes to, you know, high coupons, that could be a pro. You know, it's a pro for the investor, it's probably a con for the insurance company, you know, depending on what they're, what they're reinsuring. So, you know, I think that's the biggest that that's the biggest problem with this. And maybe, you know, the form of the question to check GPT could be slightly different where you say, you know, I'm the investor in, in in a catastrophic bond. What are the pros and cons or on the insurance company, on the issuer of a catastrophic bond, what are the pros and cons? So it's very high level doesn't really get into the weeds. And I'm not quite sure which side of the transaction this pros and cons list is representing. Most of the time I think it's representing the investor, but there are a couple of these where I think it's actually representing the insurer. So a couple things come to mind as you're speaking. One is, I think obviously it's all about the prompt, how someone asks. And so we're just asking questions that we think an investor might ask. We didn't say, hey, I'm an investor and what are the pros and cons of me buying a catastrophic catastrophe bond or a cat bond? That's one. The two is we as part of this ongoing series of man versus machine. What we've come to realize is that I would never trust ChatGPT or any of the AI machines with my personal finance question because it does get some stuff right. It does get some stuff wrong and then it leaves out some material information that would be beneficial. They do, you know, obviously ChatGPT says we get we do get things wrong consult a professional, blah, blah, blah. But if you're using it as a as a first step in learning about something, it may not be as complete as you need it to be. Yeah, I think that's accurate. I mean, just the example of are you the investor or the insurance company is a good one. I think that's an easy omission in the in the prompt, which could confuse confuse people. What's, what's your, I mean, given what you've done here and your knowledge of AII suspect you concur that you wouldn't trust any of the AI large language models with your personal finance question. Is that fair to say? I do think that's fair to say. I think it's a great copy editor, copy editor. So you could throw through some ideas to it and it might prompt you to come up with ideas for your writing. I don't think I've ever used ChatGPT word for word for anything, but it's certainly prompted some ideas. It's certainly helps me get overviews on things that may help me look into things deeper. I've in in an hour of using check GPTI feel like I've been able to get overviews of certain industries where after that I can look further into companies, you know, but but if I was going to do that, I would go, you know, straight into their company website and their actual financial filing. So, you know, I do think it could be a a start to generate ideas, but then I think the due diligence has to go much, much further than that. Well, Brian, I want to thank you for being our subject matter expert in our series, the Man versus Machine series. It's delightful for you to have to to critique chat GPD's answer. So thank you very much. Thank you, Bob.