What's The Big Deal?
Get the view from the inside. Every week, Graham Smith (ex-Ares) and Deborah Taylor (ex-Barclays) take a look at Wall Street’s headline-grabbing deals.
From mega-mergers and hostile takeovers to complex private credit transactions, they break down the why, the how, and the who behind the numbers.
What's The Big Deal?
Investment Banking: Claude Fable 5 Just One-Shotted a Bulge Bracket-Grade DCF
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
Six weeks ago, Debs and Graham asked Claude Opus to build a DCF.
The result got a B-minus from Graham and a C from Debs. Missed calculations, questionable assumptions, no clear reasoning on why it was cutting corners.
This week they ran the exact same test — same prompt, same company, no additional guidance — on Anthropic's newest model, Fable 5. The result was a step-change neither Graham or Debs expected.
In this episode, Debs and Graham walk through the Fable 5 output in detail. Mid-year discounting handled correctly from the start ("it actually knows how investment bankers think").
Terminal value presented in two methods, perpetuity growth and EBITDA multiple, side by side. Weighted average cost of capital coming in at 10.6%, in line with typical investment banking assumptions.
Diluted share count calculated correctly with buybacks accounted for. The model even self-corrected mid-build when it detected a formula error. The Lululemon test produced a striking finding: the DCF implied 45% upside to the current share price, suggesting the market may have Lululemon materially wrong.
Six weeks ago, Claude Opus solved for a value in line with the current share price. This time, Fable 5 took a genuine view — it saw upside and said so.
They also cover the broader context: CVC's Wall Street Journal-covered use of an AI agent in the Sproutz sale process, what that means for banker fees, and Graham's real-world observation that AI has moved from being a calculator to genuinely forming views on valuations.
The verdict: A-minus from Graham, B-plus from Debs.
Graham's honest closing question: at $50 in credits per model, is DIY still faster? Fable 5's fate as Anthropic transitions it to usage-only pricing is genuinely uncertain, but the six-week leap the episode captures is not.
Key Discussion Points:
The six-week transformation: from B-minus/C on Claude Opus to A-minus/B-plus on Fable 5.
What Fable 5 is: Anthropic's newest release, its release history, and its credit-limited launch.
Mid-year discounting, terminal value approaches, and WACC calculations Fable 5 handled correctly.
The self-correction moment: AI detecting and fixing its own mid-build error.
The Lululemon finding: 45% implied upside to the current share price.
AI going from calculator to view-taker: what that shift means for how models should be used.
CVC's Sproutz sale process: AI agent in the data room, and what it signals for banker fees.
The debt-treatment miss: Fable 5's blind spot on lease liabilities.
The economic question: at $50 in credits per model, is DIY still faster?
WTBD Newsletter:
https://webmail.wallstreetprep.com/whats-the-big-deal
Follow Us On Socials:
LinkedIn: https://www.linkedin.com/company/wall-street-prep/
Instagram: https://www.instagram.com/wallstreetprep/
Resources: https://linktr.ee/wallstreetprep
Alright, we now have another another anthropic model. It's literally set the world on fire, so let's see how well it does.
SPEAKER_01It actually knows how investment bankers think. I think that's really impressive.
SPEAKER_00This is quite a lot more usable than the result we got last time. Now it's six weeks. Just taking a look at this. I would say like we're in the firmly in the A minus territory here.
SPEAKER_01A minus or B plus. I mean, it's done much more complex processes than last time. It's done much more sophisticated DCF calculations. Welcome to this week's episode of What's the Big Deal? Right, Graham, what are we going to be talking about this week?
SPEAKER_00We're talking AI again. We've got a lot of a lot of news in the last couple weeks about Anthropics' recent model. I know everyone, everyone loves Fable V for coding, but we want to put Fable V to the test, at least while it's still out and available to the public, and comp it against the last Claude Excel test we did to just see how well it builds a DCF. We're going to give it a really, a really simple prompt, basically keep everything else the same and compare A versus B and see how much the world has moved on. In the last, you know, I watched our episode the other day or last night, and I think at one point we said, you know, let's let's compare contrast in say a year when the technology moves on. And here we are, literally a month, maybe six weeks later, talking about, all right, we now have another another anthropic model. It's literally set the world on fire. So let's see how well it does.
SPEAKER_01Absolutely. Yeah. So about six weeks ago, we asked Claude to build a DCF model of Lululemon with mixed results, I seem to remember. There were some a few challenges in how it was calculating free cash flow, some unusual assumptions. And yeah, I do remember you saying, Graham, let's let's revisit this, let's circle back. And I think, you know, as you said, we've now got this new version, Fable, which uh we are gonna put to the test and just we'll have a side-by-side comparison, basically, won't we? We'll be able to use the same prompt and see how the world has moved on.
SPEAKER_00Yeah, exactly. So last time I watched our I watched our last episode yesterday, and the main, the main, let's call it screw up that that I think we used Opus last time that that Opus that Opus kind of made in the in the model was it didn't even calculate no pat. And I think we we kind of we kind of agreed that in this in this case for for Lululemon, because Lululemon has zero debt, just cutting out, cutting out a couple steps on the nopad and the uh and the free cash flow calculation was kind of okay, but but also not really. It's not it's not best practice. And at no point did Claude even say, this is why I'm using this calculation. So we want to see if it's if it's grown up just a little bit.
SPEAKER_01Fantastic. Exciting times. Look forward to doing that. Um, as a bit of context, I do remember last time Lululemon had been having some challenges. Its share price had been really languishing because they had some governance issues. I know they had a proxy battle going on at the board level, and um there were concerns about the strategic direction. And actually, it seems like the world hasn't really moved on in the sense that the proxy fight has been resolved, but they're still having major issues with their strategic direction. They have a new CEO, but still the share price is languishing. So let's see what the DCF tells us about the fundamental foundation of Daily Lemon.
SPEAKER_00All right, so I'm I'm always interested to see how well these models think with really simple prompting. Now, there's a whole there's a whole science and a whole thought process behind prompt engineering. I I think the better, the better these systems get, the less, the less we're gonna have to pay attention to prompting because one of one of Fable Five's advantages as it relates to coding and software development is actually you don't need you don't need an insanely detailed prompt. It's actually pretty, pretty good at figuring out exactly what it what it is you want and what it should be doing. So I'm gonna start simple with this and see how we go. I'm I'm literally I've got I've got the the Claude Excel add in uh open over here. I've got Fable 5 selected. Hopefully I don't wanna hopefully I don't run out of credits because uh they're they're being pretty pretty stingy with Fable 5 credits these days, and I burned through quite a few of them just on other stuff. So let's see if we can get through this without without Anthropic turning us off. So all I'm gonna say is build a DCF for Lululemon. And should we just give it the the usual parameters just literally to say follow financial modeling best practices? Literally make this look like it came from a bulge bracket investment bank. That seems fair enough for me. Okay. Follow financial modeling best practices and make this look like it came from a bulge bracket investment bank. And I got a typo in there, but I'm sure I'm sure it'll figure it out. It's early in the morning here. I'm in I'm in Baltimore right now. Uh I got up at six this morning after not a whole lot of sleep, so forgive a couple typos here. All right, off to the races. So while while Claude is thinking away, uh Debs, what have you been using AI for recently in this in this kind of area? And what have been your recent successes and failures?
SPEAKER_01Oh, that's a really great question. Uh yeah, so we are, I say for a lot of our summer training, uh AI is a big theme. It's a big topic. Uh we are sprinkling AI use throughout the workflows, is how I describe it. So every key topic that we're delivering training on, we are adding an AI element to it. Um it's really important that you know, if you're a new analyst, you still need to know how to build a DCF, you still need to know how to run, you know, spread the comps uh for a company. Um, but you can definitely now automate it once you're comfortable with the fundamentals, that you can then automate it with AI. And so that's what we are doing at the moment is at the end of every key topic is showing the workflow application of AI. So the use case, uh, but then also we're spending a lot of time making sure that you know there's sense checking going on, checking the outputs like we're gonna do today, hopefully, when we've got the Lululemon DCF. Um and then also making sure that there's a really strong analyst overlay. Because fundamentally, if you're an analyst and AI is doing all this work for you, you can't just then hand that over, even if you're happy that you know it's been sense checked and it's correct, you can't just hand it over for review. There's usually some element of judgment that's needed. You need to be able to be comfortable having a conversation with a client about it. So I would argue that the professional skills become even more important. Your ability to talk to the numbers, your ability to use your judgment on top of what the on top of the calculations becomes even more important. So we're really trying to integrate that into our training. Um, in terms of the success.
SPEAKER_00I very much, very much agree with that. But why don't we let's let's come back to this in just a sec because we've got our first our first response back and first question here. Not question, rather, here's my here's my game plan. So here's my plan for a banker grade Lulu DCF. Pull the data, saying it's gonna pull data from the latest, the latest 10K. It's going to, let's see, use the 10-year treasury from official sources. Uh it's gonna put together a WAC sheet, CapM cost of equity, a DCF sheet with a bull care base case, a sorry, bear case, base case, bull case. Uh let's see, EBIT margin, DNA, capex, and organ capital, terminal growth, WAC premium, and an index-driven selected case. I'm not sure what it means by index-driven selected case. Let's see, let's see what it what it comes up with there. Uh Projection valuation. Oh, okay, five-year revenue to EBIT to NOPAT to unlevered free cash flow build, mid-year convention discounting, perpetuity growth terminal value, EV to equity bridge, implied price per share versus current. Sensitivity tables, three five by five grids at the bottom with a few different sensitivities, formatting, but that bolts bracket style, dark blue section headers, blue fonts inputs, black formulas, green sheet links, uh let's see, banker number formats, blah, blah, blah. Actually, for a pretty simple prompt, that's uh those questions aren't bad. No, Lulu has a January fiscal year end, so fiscal year 25, end of Feb 25, will actually be the last full fiscal year. I'll check if a more recent 10K ending Feb 26 is out. So it's asking now for questions. Does this look right? Anything we want change, projection horizon scenarios, or specific assumptions.
SPEAKER_01No, but I a couple of things I would just pick out of there. So this is already quite advanced because we are it's already uh it's uh identified it's gonna use mid-year discounting, which you know, when you start doing basic DCF, you assume that all the cash flows are gonna occur in one year's time. And the reality is that most bulg bracket investment banks, your mid-year discounting, because we all know that cash flows don't suddenly magically appear on the 31st of December in one year's time. They are generated over the course of a year. And so the use of mid-year discounting shows it actually knows how investment bankers think. And I think that's really impressive. It's also identified the approach that's going to be used for calculating terminal value. There are different approaches that can be used. We can use a multiple in the terminal phase, or we can use a growth perpetuity formula which has a growth assumption in it. So then I think the fact they're making these, uh it's identifying these features shows it gives me confidence that it knows what it's doing.
SPEAKER_00Do you make you make a uh a point on the terminal value? Why don't we why don't we instruct to there's two things I want to do? One, instruct to present the terminal value two ways. So let's show us show us a perpetuity growth and an EBITDA multiple, EBITDA multiple approach. Now the other, I remember the other point you were talking about last time, which Claude didn't even think about this at all, was the date from which we are performing this DCF. And you were saying in your equity research days, you used to do used to use daily discounting in essence, where I guess you've got a you've got a date that you are, but I assume a sell a sell input that you're changing basically for today, saying, give me the value of Lululemon as of this date. So do we want to instruct Claude to say, give us an input for, in essence, today's date and discount back to that period?
SPEAKER_01Absolutely. If we specify the valuation date is today, and that's the 14th of July, um, it will have to make a few adjustments. It'll have to do a time apportionment for the first year's forecast because you only got future cash flows that need to be discounted, and then we're discounting to exactly today rather than assuming a year or half a year.
SPEAKER_00Sounds good. All right, I'm just getting a little prompt together here.
SPEAKER_01Now it sounds nice and easy we're asking it to do all these things, but we're doing this because we're really comfortable that we know the process behind it. So, as you know, as we said earlier, when we're training on DCF, it's really important that you do work through the mechanics, you understand, you know, all of the different elements before you start asking AI to do all the heavy lifting for you. Because then you can sense check the outputs.
SPEAKER_00Agreed. Okay, so I just said this looks good. Two additions. One, provide two different terminal value value calculation methods, perpetuity growth and eBITL multiple. And two, provide a valuation date as of today, July 14, 2026, and allow that date input to be changed if we want to see the valuation as of a different date.
SPEAKER_01Excellent. Happy with this?
SPEAKER_00Yep. Okay. Enter. All right. Now while Fable is is building, I'll keep my eyes on it. If it comes back and has another question or two, then then we can pause and revisit. But to your to your point a few minutes ago around professional skills being a lot more important, I really, I really fundamentally agree with this. Now, I think we've probably covered this in some way, shape, or form before. But one of the things I talk to a lot of the people I'm training about is one, I think you guys are starting this work at a really interesting time in the sense that you have these tools that will do a lot of the, a lot of the initial heavy lifting, say data analysis, some of the, some of the grunt work on a much more automated basis. And your life is a bit, in some ways, a bit easier. But what it means is you're not gonna have as much experience going from start to finish, building from scratch, and knowing where one, AI can make mistakes, and two, where where you're gonna find where you're gonna find mistakes in in models just more generally. So you've got to be a lot better at looking at something and spotting trends and spotting errors and knowing what kind of questions to ask. So a lot of what I do in the classroom now is looking at model output for a model that we've built and just saying, okay, what doesn't look right here? Let's talk about this trend. Can we explain it based on something else that we know about, say the inputs to our model? Or is there an error we need to go back and check? And like that's the that's the skill set that analysts need to develop just almost immediately these days.
SPEAKER_01Yeah, I completely agree. And and it is really difficult to kind of sort of make that leap without having done all the the building yourself, as you said, that building things from start to finish can be a really edu a good education in knowing where things can go wrong, uh, knowing where things can get fudged, um, and you know, having to sort of leapfrog all that and go straight to being able to read the outputs is going to be really challenging. Um my top tip on that, and I'd be interested to hear yours, Graham, um, is to look at as many examples as possible. You know, the more that you've looked at um example DCFs, you know, if you hit the desk and you've got access to previous versions and for other deals, for example, um, or if you're in research for um other other ones built by your team, you start to get a sense of the key ratios that you expect. And that for me is really powerful. When I'm sense checking, I often look for you know what I know is a normal kind of you know level of capex to revenue, what's a normal level of working capital build, you know, the conversion of Nopat to free cash flow. You start to get a sense of those kind of numbers that you're used to. And that's where I tend to spot errors. Um but Graham, what do you think?
SPEAKER_00I think it's a combination of that. I also think one, so I was I started a new yeah, they're a mix of summer, summer analysts, summer, summer interns, or sorry, full-time analysts and summer interns. I started a new training yesterday and we were going through some basic Excel and financial statement modeling. And certainly as it relates to something like financial statement modeling, I think it's really easy as an analyst to just go through the motions, follow the instructions, and get to the right answer. I spend a lot of time talking to people about the assumptions that go into the model and say, okay, we have we have this f we have this financial statement model, we have this set of inputs and assumption drivers. I want us to figure out which of these, which of these is the most important. And when we when we look at some of the model outputs, really try to explain trends based on based on this set of inputs that we have. There's, I'm sure you probably use the same example. Like we use a an old, I say old, it's probably five years, five years out of date, because date dates from this perspective don't really matter, right? It's more, it's more about the the actual mechanics of putting something together. But it's an Apple financial statement model as of 20, I think 2018 is the last, the last actual forecast year, or sorry, the last actual historic year. And then 2019 forecast a revenue decline. 2020 has revenue going back up. So because in an FSM model, as an example, you have so many things tied to revenue growth, you get some funky looking inversions. You have some working capital unwind in 2019 that then reverses in 2020. So I'm always making sure as we're going through to say, okay, what about this doesn't look right? Or what about this looks funny? And then based on our assumptions, does this make sense?
SPEAKER_01That's a really good point. And I wonder if we might spot some of that with Lululemon because as we said, it's been going through a bit of a turnaround. And I think that does make it quite challenging. What we tend to like, well not like, but what we expect with DCF and forecasting in general is nice, smooth, gradual growth. Um, and that's certainly not what we would expect for Lululemon. Um how's how's Fable getting on?
SPEAKER_00Okay. So we've got we've got another another stop here now where it says, okay, here's the data I've pulled. What do we think? So it pulled a few, a few years of actual results from from 10Ks. It's got the share price, it's got diluted, diluted shares outstanding. It looks like it's done, okay, diluted shares outstanding, 114 million. Let's see, 111 million common plus 5.1 million exchangeable as of, let's see, at fiscal year end per the 10K, less 2.2 million quarter one buybacks per the Q1 release. Okay, this is actually this is actually done a decent amount of digging in terms of just just news to say, okay, what is the right, what is the right diluted share count here? Okay, net cash, zero debt, undrawn $600 million revolver. I think we knew that from from last time. So that that looks like it's remained constant. Tenure treasury, 4.58%, beta assumption 1.2. It says no official source publishes beta, blue input that you can flex. So it's being honest about the fact that it's just made an assumption here. Tax rate, 30% management guidance. Okay. Then we say, okay, context, 2026 guidance was cut in June. See, we've got the, we don't need to run through all the numbers here, but we've got the the revenue growth uh and margins for the bear case, the base case, bull case, all cases, DNA, 4.5% of revenue, capex going from 6% to 5. Let's see, it hasn't hasn't specified why it's making that assumption, but I'm sure we can ask the question. And the whack the whack at 10.6%. Valuation date input, July 14th, 2026. Drives all mid-year discount periods dynamically. Terminal value, perpetuity growth, and even dumb multiple approach side by side with a method toggle feeding the headline bridge.
SPEAKER_01This this looks a lot more robust. I seem to remember last time the whack was a bit odd, wasn't it? For the um for the the original DCF that was built. So 10%, that's kind of my go-to starting point. Um so this is looking sensible. And even the CapEx, I think last time it gave a really punchy CapEx assumption. Um so yeah, let's see, uh let's see what the outputs are then.
SPEAKER_00I just said go ahead and build it. So let's uh let's let it crank over here. If it if it stops and asks another question, we'll we'll take a pause. Otherwise, we'll we'll see what it comes up with when it when it finishes building.
SPEAKER_01So, Graham, while whilst it's doing that, I've got a question for you because uh when you said you wanted to put Fable to the test, I will confess, I wasn't really familiar with Fable. I haven't used it at all. Um and then you mentioned just earlier today about the fact that it's kind of allowing some free access. Do you know the situation there? Is this kind of like a bit of a teaser to kind of get people locked in?
SPEAKER_00That, you know, that that's a good question. I mean, so the I don't know if you know the backstory behind this model, but you I I'm sure people have probably heard of the anthropic mythos model, the thing that was basically so good they had to withhold from public release. Yeah. Fable is the more general release version of it. I think it's basically the way I understand it is it's the same model, but without some of the crazy cybersecurity kind of hacking logic that people use mythos to bolster their security with. That's the one they really don't want to get in the wrong hands. So they released Fable V, I want to say about a month ago, thereabouts. And then pretty quickly thereafter, the US Department of Defense said, hey, this is so good, you can't put this in the hands of anyone. Only US citizens should be allowed to have access to this, to this model. And really the subtext there, I think, is it's basically the US government saying, at this point, we're picking OpenAI and Sam Altman as the winner in this space. I'm not, I'm not sure, I'm not sure it was really motivated by anything other than just political, political goals. Long story short, they they as an anthropic came to an agreement with the US DOD to get Fable released again. But to the to the sort of disappointment of a lot of normal users, I say normal users, I mean I I have the like literally the $200 a month Claude subscription. So it's not like I pay nothing for this, but Anthropic has said, we're gonna give you Fable V. It uses a lot more usage credits, you only have so much access to it, and after a certain date, it goes to usage only. So it's not included in your plan credit anymore. You have to basically put the quarters in the machine every time you want to literally get some some Fable V tokens. Now, I don't know where where this ultimately goes because it's kind of funny. When you log, I mean, let's see if if it even shows up in this interface here. No, it doesn't think it's probably in the in the normal cloud app, but whenever you go to choose the model, it says Fable 5, include it in your plan until July 7th, then July 12th, then July 14th, then it's July 19th. So I don't know, I don't know if they're if they're just pushing it back to get as many people addicted to it as they possibly can, and then they make you pay per use. I have to think at some point though, I mean, ultimately the world is competitive. OpenAI will come out with something, with something good, and then Fable will get included in your plan usage, but then it's gonna get superseded by the next latest and greatest model. So it's always this little chicken and egg situation. But I have to assume the reason they're doing this is to try to maximize their run rate revenue in advance of the IPO.
SPEAKER_01Yeah. Okay. Yeah. That's it, that's a good point. Um, but it is interesting, isn't it? That actually we started off, all of us I think started using um AI with no real concern about limits and even the really the cost of it, it seemed, you know, something that you could use in a very affordable way. And already so many of us at work are finding, you know, we have subscriptions, but you know, on a daily basis, you reach your limits, you know, and then you have to request more tokens. And you can see at some point we do reach that point where you're not having to make an economic decision around the cost of using you know AI versus someone actually having to go through the thought posts themselves. And so far it's been so much cheaper. But you know, there will come a point where you're you're thinking, hang on, there's an analyst that you can pay 100 grand a year versus you know a number of tokens which you know are creeping up in price.
SPEAKER_00100%. And the other thing is to the note or to the point of having to know how to do this stuff yourself. What happens when your tokens run out? You're like, oh wait, I've never done this before. I don't know what I'm doing. I just have to stop working. No, you still you still need to know what you're doing.
SPEAKER_01Yeah, absolutely. But how is it on?
SPEAKER_00It's it's it's just churning. So the one, not the one, but one of the one of the main distinctions between Fable and some of the other models is it will just crack on and get on with it. It doesn't, it doesn't necessarily stop and ask as many questions. We had the two, the two kind of question rounds. Is this approach right? Okay, here's the data. Am I okay to go? Now I'm very much expecting the next the next step here to be pretty pretty fully thought out, unless there's unless there's a major question or something happens along the way and it says, hey, what do you think? What do you think about this? So let's give it, let's give it a few minutes and see and see what it'll be pulls pulls together. But it looks like if I if I look through the the steps it's following, says setting up sheets, build the WAC, controls and market data, build scenario assumption blocks, build financial projections in DCF, build discounting, terminal evaluation, and valuation bridge, build three sensitivity tables, format, and then verify. Doesn't sound like a crazy, a crazy list of steps. So let's see how it goes.
SPEAKER_01So, Graham, uh, we're demonstrating um a really clear, obvious use of AI in terms of building a DCF, but we've been reading in the news, haven't we, about other tasks which um AI is being used to kind of replace some of the workflows that investment bankers traditionally do. Can you tell us a bit about um one of the deals that you were highlighting to me earlier?
SPEAKER_00Yeah, we just saw, I think it's in the last week or so, there's uh there was a story in the journal about CVC using an AI agent. I mean, this is an attention-grabby headline, right? It says C C VC uses AI to replace investment bankers to run a sale process. And it's that, but it's also it's also not that. And basically CVC was selling, selling a portfolio company, it's called Sprouts, and it was the article was touting the fact that in the data room was an AI agent that was there to ask prospective or to answer prospective prospective buyers' questions. Usually you'd have a team of investment bankers and analysts really prepped to go through the data room and come back with any kind of with any kind of questions. Here we've got an article, article basically saying this is the first time that AI has been used to do this. Actually, I'm I would be I would be shocked and be 100% shocked to to to find out that this is the first time this has actually been the case. Because if you're going through a data room these days with the with access to AI, are you using these tools to pour through the data room and answer the questions you have? 100% you are. I think this is the first time where the seller or the banker, at least publicly in the news, has said, hey, it's part of our data room. We're in essence including an AI chatbot that will help you go through all the data and answer any questions you might have. And by the way, I'm sure we've talked about it before, but I think that that skill in particular is one of the best uses of AI today. Just going through these large data sets, answering questions, pulling out trends. I mean, in the in the FSM modeling exercise, I was running through with these, you know, first year, first-year analysts and summer associates yesterday, part of the exercise historically was going to public company filings, going through 10Ks, pulling out data. And I kind of said, you know, I'm not gonna, I'm not gonna make you guys do this. I mean, not not least because you guys are in mid-market private equity, you're not gonna have to go through public company filings, do diluted share count calculations, all this kind of stuff when you're when you're building a model. Uh, but one of the things I talk about is just how good AI is if I take that model template and say, okay, Claude GPT, go pull out Apple's last three years of historic filings from literally back to 2018, fill out the inputs for the sheet, pay attention to the notes and some of some of the columns saying, okay, grab, grab this piece of data from this page. And it will literally fill out the entire input sheet. We'll have a comment in every cell saying, This is where I got this number from, this page of the 10K or the 10Q. For that, for that kind of work, it's really, really impressive and saves a ton of time. So it doesn't surprise me that we're seeing this being the first public use of AI that we're talking about in the news. But no doubt, privately, this has been the case for quite some time already.
SPEAKER_01Okay, but it is kind of a step on the journey to some of the work that's previously been done by investment bankers, being done by AI. Um and that is potentially a risk for you know some of the fees that are being generated by the investment banks for their sell-side process, isn't it? Because you know, sell-side process has a number of steps. You've got to think about, you know, the marketing that you're doing, trying to generate interest in the buyers, managing the data room. And as you say, part of that is now, you know, the you know, answering questions from prospective buyers. There's still other stuff that's you know, probably still being done by um the actual bankers themselves, things like advising on the valuation, you know, we've talked about things like valuation opinions and things. So those are still, you know, there might be sort of small cogs in the process which are gradually being transferred to AI. Um, but it is a risk, isn't it? That you know, the bankers' fees start to go down if more and more of it can be done by AI, and specifically that more of it can be done by the firms themselves rather than the advisors.
SPEAKER_00Ooh, when do you think we're gonna have the first AI signed fairness opinion where Claude says, I think this deal's good?
SPEAKER_01That is a really good question.
SPEAKER_00Or fair enough.
SPEAKER_01Yeah, it it's about it's about risk, isn't it, really? Because the fairness opinion is you know, the actual work involved, as we've we saw when we looked at the um it was Paramount Sky Dance Warner Brothers deal, that the actual work involved was looks quite limited, but it's the risk that you're taking on because you can get food if you provide opinion that doesn't hold water.
SPEAKER_00Um Yeah, no, 100%.
SPEAKER_01Interesting. Right, how's how's how's the uh model getting on?
SPEAKER_00We let's see. It is, it keeps building. It said isolating which formula failed. So it's made, I don't know exactly exactly what it what it failed at, but it highlighted that it made a mistake. So now it's now it's going through and correcting itself. It's still it's still crunching. The failure was the LET formulas. What what's L E T, Debs? Do you have any idea?
SPEAKER_01Absolutely no idea. It well, it says it's not supported in this Excel build.
SPEAKER_00So I'm in a kid with Excel. Is it hallucinating some some finance acronyms now?
SPEAKER_01I mean it is quite good that it's sort of self-corrects, you know, for for a long time. It was kind of as you say, it was kind of like really needed a little bit of coaxing, didn't it? Come back to you and say, Oh, I found a problem, what should I do?
SPEAKER_00No, a hundred percent. It's just it's just getting on with it. It's just getting on with it. I'm very much expecting by the time it finishes getting on with it, we'll have something that's at least at least semi-usable. Am I expecting this to be perfect and as good as a really good analyst if they've gone through and done the job properly? No, no, probably not. But on the basis of some really, some really quick prompting, I'm expecting we've got something at least semi-usable. So, Greg, we'll give it a few more minutes here.
SPEAKER_01I can see some assumptions though on the DCF. Should we have a quick look at the assumptions on the DCF tab and see if there's anything that jumps out at us?
SPEAKER_00Let's see. Okay, I was gonna I didn't want to interrupt it too much. I was gonna I was gonna resize some columns just to just to let us see everything, but it's already it's already gone through and and done some of that. Okay, so we've got okay, we've got a case, a case selector. Let's see. We've got a choose function, not necessarily my favorite, my favorite function for this, but you know what? Not not horrible. Not horrible. Okay, terminal value method. We've got, okay, one perpetuity, perpetuity growth, two EBITDA, EBITA multiple approach, current share price, dilute shares outstanding, market cap, cash and cash equivalence. We we already know we have no debt. So we have negative net debt, enterprise value, and whack of 10, 10.6%, which I think is about what it said it was gonna do in its setup assumptions. And we'll go through and we can take a look at the whack calculation in a second. Okay, so we've got three three cases here: bare case, base case, bull case. Bear case looks like we've got revenue down 2% this year, flat next year, and then 1% growth 2028, base case down half a percent, 3% growth and 4.5% growth, and then bull case half percent growth, 5% growth in 27, and 7% growth in 28. You know, I I don't know the company that that well, but directionally, just in terms of thinking about what we know about historic results, what management guidance is for the next year, just looking at say the variance in these cases based on the number of years, going there's not there's not a huge range in the the bear case to the bull case this year, right? And even in the even in the bull case, revenue growth in 26 is only is only half percent, right? And then and then as we get in the outer years, it's making some slightly some slightly divergent projections about what those three cases are going to deliver. I think that that displays at least a bit of thought that is above and beyond just simply saying my bear case is negative 5% every year, my base case is flat, and my bull case is five percent growth every year.
SPEAKER_01Absolutely. And I've got some numbers in front of me um for consensus, and it looks like those base case numbers are pretty much in line with consensus. It's found some uh numbers somewhere which are publicly available showing consensus information. And that but yeah, the starting point usually for bull and bear cases, you're you're going for like 100 to 200 basis points, that's 1% or 2% above and below uh the base case in terms of revenue growth and then an adjustment to margins based on you know what is reasonable within the industry. Um, but yeah, it looks it it looks like a pretty sensible profile across all of the cases. Um and also, as is normal, we only sensitise revenue growth and margins. We don't sensitize usually um the other assumptions unless there is something you know really specific that you know could happen in terms of store growth, store expansion that affects only one of the cases. So um all the focus of the bear case and bull case is around growth and margins, which is yeah, sensible.
SPEAKER_00And by the way, I just I just hid the window because it said the model the model is complete as described. So now we've got something at least at this at this stage to go to go and take a look at. So let's I actually just want to see also just how it's how it's been built the the model. We got a selected case assumptions. This drives the model. We're using some choose functions to pick between those three cases above. Again, for a for a model with this with this number of inputs, i.e. not that many, it's not a not a horrible way to go about it. I'm not I'm not super mad about this. Okay, and for each, we've got a we've got an input for terminal growth rate and exit multiple assumptions. So on our bull case, we've got a 9x exit multiple, 8x, and 7x. And then for our for perpetuity growth on our base, or sorry, bare case, we've got 2.5%, then three, 3.5% growth on on perpetuity. Okay. Doesn't sound doesn't sound crazy. All right, now let's look at our let's look at our unlevered free cash flow calculation because this is what it really it really did not think about or think about properly last time. We've got three years of historic results. Just want to check these, uh check these comments here. All right, it's saying exactly where it pulled each one of these figures from. Then we've got a uh we've got a calculation for for revenue growth, which is assuming is just going up and pulling from our case assumptions up above. We've got EBIT driven by an EBIT margin. We've got depreciation amortization, depreciation amortization. Let's see how we're calculating this. We are just taking a, we've got an assumption for percentage of revenue. Again, for a for a high-level model like this, I don't think that's crazy. Right. Gets us our EBITDA. Let's see. Then we've got, I mean, it's slightly funny presentation, but EBITDA and then less taxes on EBIT. It looks like this calculation is is actually correct. And then, and then no pat. So EBITDA is just a is just a presentation line here. It does look like it's actually calculating no pat correctly. It's taking EBIT less taxes on EBIT. Then add back DNA, subtract capex, changes in networking capital. So it's got it's got an assumption for for networking capital, which is not, let's see, it's not, it's not a huge networking capital impact. And actually, how how have we thought about that assumption there? Again, okay, just networking capital balances, percentage of revenue. Again, by the way, how do you how do you forecast networking capital balances when you're doing a high-level model? Do you pick a percentage of revenue and forecast that way? 100%.
SPEAKER_01I think a couple of things, a couple of sense checks that I always run, capex ahead of DNA, usually, because you need a company to spend more on its uh on its growth uh than the depreciation on its existing business. So you've got a nice ratio of capex to DNA there.
SPEAKER_00Um the network one thing, by the way, on that, what I don't know is based on if we gone through and really, really said, okay, what is what is Lululemon's expansion plan for the next couple of years, would you make some specific assumptions here? Yes, this this looks a bit more high-level, but for for this stage, I think that's probably okay. Absolutely. Sorry, go ahead.
SPEAKER_01Um and then the other thing, it's something that you mentioned um that is to do with the networking capital, that sometimes when you've got a contraction in revenues, the fact that you can end up with some slightly uh wacky working capital flows. Um, because usually a contraction in revenues, you know, you know, if a company is struggling, it's not usually going to result in cash generation from working capital. If you link working capital to revenue, you can end up with this odd working capital inflow. Um, they've got a very small three million dollars or whatever. So um it's a small inflow, but it's actually not too crazy, um, which again is quite really short.
SPEAKER_00Okay, let's look at the actual, let's see, discounted cash flows, fiscal year end date. We've got a, let's see, year ends. This is in essence just running running a year frack formula here. Let's see. I'll never free free cash flow included. What's it okay, what's it doing here? Fiscal year end date.
SPEAKER_01So that's a bit weird because I would usually expect it to be the year fraction to be one for every year except the first forecast year. Because what you're doing, if you're discounting to July and you've got a January year end, you've got a certain number of months until the first year end, and thereafter we're going to be discounting full years. So I'm not quite sure why you end up with more than the year.
SPEAKER_00Okay, I think what we're in essence, in essence it is that, but it's looking at the fiscal year end date, and it looks like some years fiscal year end is January 28th. Oh, it's a 52 week. February, yeah. Yeah, okay. Okay. So that that I think is the the the issue here. Again, it's not this isn't gonna this isn't gonna change the math materially at all. Okay, then we've got our our discount, our mid-year discount period, discount factor. Let's see, where's our where's our discount right here? Okay, weighted average cost of capital, okay. Present value of unload free cash flows. Then let's look at our terminal value calculations here. Two different two different approaches. Terminal, terminal, terminal year free cash flow, one plus one plus the growth rate. We're using the the normal uh perpetuity discount uh discount formula here. We've got an implied implied EV to EBITDA multiple based on that perpetuity growth rate calculation, method two, eBa. Okay. You know, it's not crazy.
SPEAKER_01It's not crazy, is it? I mean, even just seeing the implied multiple, high single digits, low double digits is kind of the rule of thumb, isn't it, for a terminal multiple? Um terminal value as a percentage of enterprise value, 73.5%. That's a nice little sense check in there. Make sure you've not got all of the value baked into that terminal value.
SPEAKER_00And by the way, we didn't even prompt to say give us that. It's just we, I mean, the prompt we said was follow Bulge Brack and Investment Banking best practices. And it has just figured that out. It's just made a decision. Hey, I might, I assume they might and they might want to see that. Right? Actual valuation summary. Okay, enterprise value less net debt equals equity value. Net debt is negative, so equity value is higher than their enterprise value. Okay, diluted shares outstanding, implied share price, current share price, implied upside 45%.
SPEAKER_01Oh, yeah. So this is the crux of it, really, isn't it? So basically, Lulu Lehman as a company is going through some really challenging times. We mentioned the share price is really low at the moment compared to, you know, even a year ago. And the DCF is still showing um a fundamental value for the company, which is well above the share price. That's really interesting, isn't it?
SPEAKER_00Now, from memory last time we did this, Claude basically solved for a current share price for evaluation that was basically in line with the current share price. Here it's here it's actually taking a view. It's saying, hey, I actually think there's there's upside to the share price here. Now, the one the one thing that it's still, from my from my perspective anyway, I am this this format, I think, is technically correct. I mean, I would anytime we're doing a sensitivity analysis like this, I would always just use a data table because it's a lot easier to figure out what's going on. I'm sure we could go back and prompt Claude and just say, hey, these sensitivity tables use an Excel data table instead of this crazy some products, like whatever's whatever's going on here. I just like, I can't, I can't audit that formula. I don't know about you. I don't know if you can look at that devs and just say, yeah, that's right.
SPEAKER_01Absolutely no way. But I think what's really interesting is that clearly AI, the default approach now is to make everything as auditable as possible, which in a sense is good. It's just you do end up with these ridiculously convoluted formulas, which in theory are auditable. But to the average analyst, because you never build these formulas yourself, you as you say, you just run the data table, which is basically little macro, and all the outputs are hard-coded, we would never sort of know whether those formulas are correct or not.
SPEAKER_00So um Exactly, exactly. But what is what is important is can you at least can you identify trends and and decide whether these are directionally correct. Absolutely. Right? Because here we've got we've got our our share price calculation sensitized by the WAC and terminal value growth rate, right? Share price going up as we increase the growth rate, going down as we increase the the cost of capital. Same thing with revenue growth and margin, beta risk-free rate. Right. Can you can you take the can you take the inputs and figure out do you already know what it's supposed to do in terms of affecting the outputs and at least look at something and know if the trend is right? That's that's your first step. Right. Would I would I go back and prompt to replace these with data tables just to make it really easy? Yes. But as a as a first stab, you know, it's actually pretty good. And if you have Excel set to partial calculations, this is still gonna calculate every time without having to hit F9. So I guess a little a little bonus there. We are getting, we're running, running close on time for today. I know we've both got to get out and teach and get into the classroom, but we talked last time and I said within a couple years, we're gonna get to the point where AI is gonna be, at least for this kind of work, as good as that first year analyst. Based on this prompt, build ADCF for Lululemon, follow investment banking best practices. And fine, we gave it some specific instructions in terms of hey, provide the the terminal value on two different two different methods and give us a date that we can that we can use for the discounting. That was it. Yeah. That was literally all we gave it. And this is quite a lot more usable than the result we got last time. And that was six weeks.
SPEAKER_01So, Greg, am I right in thinking that last time we graded it and we basically were pretty down on the results?
SPEAKER_00We gave I think I think I I gave it, I gave it a B minus, and I think that was being generous. I think you gave it a C C something.
SPEAKER_01So what would you give it this time, Greg?
SPEAKER_00Ooh, ooh, ooh, ooh. Okay, based on in terms of in terms of output per per prompts, I mean to get the full grade, I'd have to go through it, I have to go through all the source filings. I'd have to read the press releases and see, you know, has it has it actually has it actually parsed all that stuff correctly? Has it come up with the right view on bear case, base case, bull case? Like that's that's the the real detail work you need to properly grade it. But at a high level, just taking a look at this, I would say uh like we're in the firmly in the A minus territory here, I think. I'd say I would say time for time spent.
SPEAKER_01A minus or B plus. I mean, it's done much more complex processes than last time. It's done much more sophisticated DCF calculations. There's one little niggle I have, and it's something I didn't pick out last time, which is um on the debt side, it treats debt as zero. The reality is they've got a huge store portfolio, which is leased. Technically, I would treat that as debt-like. Um, it hasn't flagged that. So I would want to get, you know, roll up my sleeves and sort of dig through and just check that that isn't creating an actual error in the output. Um, but I think, I mean, I'm I'm warming to this now. I reckon B from my perspective, which as you say, in six weeks is a massive improvement.
SPEAKER_00Yeah, that in that in that short amount of time. So should we do this in another six weeks and see where we are? Hey, and you know, the one yeah, we haven't even tried, we haven't even tried GPT for this. I think generally speaking, I've had better luck with the Claude Excel Excel plugin. Uh so we can we can pitch them against each other. Maybe we'll do a side by side, give them both the same prompt and see how they do. But you know, in in not that much time, some pretty pretty impressive progress. So if Able 5 does seem to be living up to its reputation, do people continue paying for it when you've got to buy the credits and it costs $50 to put this model together? Would I guess go through and just do it myself? Like maybe. You know, that's the that's the question to answer the next the next like few weeks or so. Absolutely. But we'll see, we'll see what anthropic decides to do about it.
SPEAKER_01Great. Well, um, thanks thanks to those of you that listened to our uh our little challenge for Fable, uh building a DCF for Lululemon. I hope you found this episode interesting. That's all we've got time for this week. But uh that's thanks for me and see you soon.
SPEAKER_00All right, thanks, Debs. And by the way, if you want to see us do any more of these or if you've got a modeling challenge for either Fable or the latest OpenAI model, let us know in the comments down below and we'll try to get to it in an upcoming episode. But until then, take care, everyone, and we'll see you same time next week.