In the middle of my 10th year, it has been a while since I wrote Reflections from Year 4. For this one, I tried to think of what is discussed most often on here and provide thoughts on a few of those topics. As always, written with the target audience being my younger self.
Is Asset Management Dying?
Flows have been tough for the industry, but I will tell you that fees will always accrue to performance and there will always be a market for funds that outperform. We have all seen the studies showing the majority of mutual funds underperforming net of fees, and while I would encourage you to check sample size/quality in these studies, I actually do agree with the premise that most funds are pretty crap. I’ve met enough people in this industry at this point, a lot of them with extremely impressive resumes that blow my own out of the water, and I am both typically unimpressed and encouraged by being unimpressed. I am unimpressed because it is extremely rare that I hear a differentiated viewpoint on really anything, and encouraged as I think surely I can do better than this guy. Perhaps that is arrogance, but you have to be a little arrogant to come into this industry and compete against all this highly paid ivy league firepower and think you can do better despite the industry’s aggregate crap track record. So, I would say that the crap funds are dying, and should die, but they are dying a death of 1,000 cuts and someone asks if the industry is dying every single year that I have been on here. My view is that there will always be a value attached to risk-adjusted outperformance, and that it is possible to generate said performance through-cycle.
Taxes & Market Appreciation
So why haven’t the bad funds died off faster? One thing that doesn’t get a lot of attention is the impact of capital gains on attrition. Particularly in the HNW channel, people hate paying taxes. If you put a client in a mutual fund in the late 90’s or early 00’s you now have >20Y+ of embedded capital gains in that fund. If the fund’s returns haven’t been a complete disaster (meaning in the neighborhood of the bench net of fees), it doesn’t make a ton of sense to churn that client into a passive vehicle or some other fund at this point and stick them with that big tax bill as the tax bill may outweigh the savings on fund fees. This is a different calculus for institutional investors who can be more creative with their tax situations, but for individuals and the wealth management channel, this slows the rate of withdrawals and churn.
The second thing is that a shop that has underperformed over this time needs only the market to outpace the rate of withdrawals in order to manufacture fee growth. The S&P 500 has been compounding north of +10% for the last 20Y. As long as you are losing money out the door at a slower rate than that you have been growing your business. So, you may have created absolutely no risk-adjusted value net of fees and not had a positive flow in 20Y and still your shop grew just because of market appreciation. A real, multi-year bear market might hurt a lot of folks but this might actually result in creative destruction and be a longer-term positive.
Active vs Passive
This also has been asked every year that I have been in the industry. My view continues to be that there is an equilibrium, but it is impossible to say what the right balance is. Perhaps we’ve already gone too far passive and that will create an opportunity for active to generate alpha. But what I always like to say on this topic is imagine a world that is 100% passively invested. In this world, there is no price discovery, and price fluctuations are purely dependent on asset flows in and out of the asset class. Additionally, in this world, capital is free. Companies can essentially do whatever they want because stock prices are no longer reactive to changes in fundamentals as no one is doing the work. It would also completely violate the EMH principles on which passive investing is based. The logic is circular – you can’t beat the market because prices reflect all available information, so just invest passively, but for this to be true someone must be incorporating all available information by actively investing on said information and thus incorporating it into stock prices via that trading activity. The very act of someone incorporating information into stock prices by trading activity is irrational in this world because they should know they can’t beat the market.
Is AI going to Replace Analysts?
This is a topic where what I write here may not age very well given how quickly things are evolving. I would tell you that I have been extremely impressed by how far and fast the technology has come, and I am now a daily active user of a couple models. I have two primary thoughts on this topic. First, these tools are only as good as your ability to provide creative and useful context. If you ask simple questions you will get simple answers of little utility. If you can craft a creative and detailed prompt, you will get a deeply useful answer; however, I would posit that to even be able to ask the right questions, you have to still do the same work and research you had to do before. You have to know what you don’t know, and you can only identify holes in your knowledge base by first forming said knowledge base. The models still cannot do the work for you. Instead, I believe the models allow you to go deeper, faster, and they help with organizing thoughts when it comes time to communicate the idea to others.
My second thought applies broadly across industries, not just finance. Many of the initial AI use cases have been aimed at replacing junior positions and/or steppingstone-type positions. But if you eliminate more junior roles you eliminate your future funnel of talent. This is true of every industry. I don’t think you can broadly eliminate these positions if you intend to staff your more important senior roles several years from now, and I think eventually this will become evident.
Training your Human LLM
You really have to invest in yourself to remain relevant for the long-term, and that means continuous learning. I personally attribute most of my ability to break into this industry and remain here to independently studying and learning. As you do the research, you need to have the information bounce around in your neural net and have it be appropriately weighted as it filters into a useful output. You need to train your model, and while there is no real substitute for on-the-job experience, the easiest and most accessible way to bend your personal development curve is by reading/learning about a variety of directly and indirectly related topics. Train your model so you can get better at inference. If I could give you one piece of advice to stick from this section of this post, never think you have it all figured out. I think one of the reasons that people suck at this job is because they think that breaking in is validation enough and that once you are in you’re done. See the Dunning-Kruger effect. Never stop learning and never think you have ‘it’ figured out. Education, particularly for this industry, is a journey not a destination.
The answer is both, not either or. Look, I get that the CFA is a grind and that MBAs are expensive, but there is ROI on both. Why do the CFA? The CFA is not a golden ticket to the industry, nor does it signal a readiness to do analyst work. What it does signal is work ethic and commitment to the career path. The MBA is much closer to a golden ticket, particularly a top 20 MBA, and it gives you significantly more career optionality, but it is pay to play. The CFA can be done in your free time for little financial cost, so there is relatively low risk in just getting it done or even knocking out a level, particularly early on in one’s career when you arguably have the most free time. The MBA is a bigger commitment, but I do think it is worth it. And if I am wrong about the above RE: death of active management, the MBA gives you some downside protection via the alumni network and ability to craft a story for switching industries. So, when I do networking calls with recent grads that want to break in, I always say do the CFA and in a couple years (say once you have 5Y of work experience) if you haven’t broken in go crank out a good GMAT score and get your MBA from a top-20. This is the only proven high probability path to breaking in that I know of.
Make sure that during the interview process for any shop you ask about the makeup of AUM and understand how concentrated the asset base is. AUM per head might be good, but if you lose a couple of large clients and the horse gets shot out from under you that safety will prove to be illusory. Also, related to the above, don’t sleep on shops with a lot of PWM money. I used to think that would be a negative as retail money is most prone to panicking, but it is much stickier capital than I thought. While institutional capital can scale much more quickly, that goes both ways, and I’ve talked to many in this industry where their shop lost a big institutional account and then like a vortex other institutional money followed out the door very quickly. As an investor, I am always acutely focused on downside scenarios, and when you’re searching for a shop you should be thinking through the range of outcomes the same way as you would any investment.
Alpha vs Beta
I wrote about this in the 4Y post, and it is still as important as ever. I think most long-onlys are simply riding beta or some singular style factor. This is fine when timed correctly and can have very positive results for as long as you have momentum at your back, but it is not a recipe for longevity if it is done so blindly. That said, while the name of the game is generating risk-adjusted alpha, managing and correctly timing beta is equally as important and valuable as a skill. There are many who say they don’t ‘time the market’ or that they don’t do macro, but this is silly IMO. You can still be a disciplined investor committed to a particular philosophy that is primarily bottom-up focused while managing/thinking about your top-down risks and exposures. The long-only game is difficult enough searching for alpha while trying not to get run over by factor-beta moves, but you have to know what bets you’re putting on as a starting point. Related to this, beware of the funds that are consistently in the 90th or 10th percentile any given month or quarter. This is a signal that they are over-levered to a particular big idea. I like this quote from Taleb that I have posted before: “at any point in time, the richest traders are often the worst traders. This, I will call the cross-sectional problem: At a given time in the market, the most successful traders are likely to be those that are best fit to the latest cycle.” Also see the concept of strategic mediocrity.
I am Still Loyal to the DCF
I’ve been a supporter of the DCF approach for a while, and my conviction in this approach has only become stronger with time. It is probably the most underrated tool in all of finance. Stock prices are simply math problems. We can argue about the inputs and qualitative narratives, but really how that all comes together to form a stock price is a matter of arithmetic, not opinion. I think there are many out there who believe multiples are a form of mystical market magic. How many times have you heard someone say “that’s just how it trades”. Now, that isn’t to say I don’t look at multiples, I do. But you have to have a handle on 1.) the math that underlies multiples; and 2.) if you are comparing a current multiple to a historical range, the numbers that were underlying those multiples earlier in the company’s history. You cannot simply look at a chart and say a stock is cheap or expensive because it is north or south of some line drawn on a chart. A multiple is a direct function of expectations for future ROIC, growth, reinvestment, and risk. This is what makes comps analysis dangerous – you can’t really compare two stocks unless all of those variables are fairly similar, and even then I only bring comps analysis in as a tertiary support for a thesis. Also not a sum-of-the-parts guy, these are often highly flawed when I see them in the wild for the same reasons.
The sell side is not the buy side. Buyside investors allocate capital to make money. Sell side analysts have a different set of incentives, and they allocate no capital to their ideas, thus, it is not their views that are reflected in stock prices and it is important to understand this when thinking about “consensus” estimates. Garbage in garbage out applies as much to multiples as it does to a DCF.
First, the largest bank analysts have very little incentive to make bold calls and will be the least divergent from center most of the time – if you’re already a career analyst at JPM you’re already in and probably pretty comfortable. The smaller shops are much more likely to espouse more extreme views or estimates as no one is reading their research anyways if they are wrong, and if they get a big call right they could end up on CNBC and make a name for themselves. So understand where an analyst sits on that spectrum when thinking about what they are saying/writing. Second, there’s a reason that almost everything is buy-rated and a hold is basically equivalent to a sell. The sell side largely understands that management access is their primary value-add, and if they piss off a management team their access is going to get throttled. As a result, consensus estimates in the NTM are simply management’s guidance as no one wants to piss them off, and further out estimates have a bullish skew to them for the same reason. Third, the sell side knows who is paying their bills, and it is not lower turnover investors. They get paid off trading commissions, thus, the bulk of the research is shorter-term oriented and aimed at funds with higher turnover. As an extension, these are the funds they talk with most frequently as the greater commission dollars affords those funds better analyst access. This is probably not the game you are playing at most long-onlys, so realize you probably aren’t the target audience for their work and your mental models for evaluating an opportunity probably aren’t well-aligned.
The Singular Biggest Differentiator
I think the singular biggest differentiator between a good investor and a bad investor is self-awareness. I do think that this is partially innate – for someone lacking self-awareness, you’re not even aware of your deficiency and cannot take steps to address, a bit of a paradox – but it can be learned, developed, and improved upon over time. IMO lack of self-awareness is the root of all investment biases, and biases are the cardinal sin of the investor. Everyone in this game is well-credentialed and intelligent, but it matters not what you think, but how you think. You can be both intelligent and emotional. The best investors I have met are deeply introspective, intellectually honest, and a bit self-conscious. It can come off as a lack of confidence, but this again is a bit of the Dunning-Kruger effect. They know enough to know how little they know, which is a direct function of self-awareness IMO.
Great write up, +1 SB
That tax component is such a good point that I've never considered at all.
I have been using an analogy for the active vs. passive that you highlight and I think it captures it well in layman's terms....imagine you are on one of the pedal hopper bikes with 5 of your friends....if one of your friends stops pedaling, not a big deal. if 2 stop pedaling, it's a little harder, but not unmanageable. but if all 6 of you quit pedaling, you're not going anywhere. and that's the rub with passive...someone has to keep pedaling for it to work.
I like that, that's a good way to articulate it
Idk if the question is as much if AI replacing analysts, as it is replacing the associates underneath them. They do the grunt work that can/will be replaced. AI can likely write the notes better, soon make the models better, and can take in all types of news faster.
There were a lot of good points you made. (edit: This) is a good, genuine value-add post.
(1) "I am Still Loyal to the DCF... Stock prices are simply math problems."
But my god. This quote is very funny in how definitionally incorrect it is. I get the point you were making here, but I know you know this isn't how stock prices work.
It is certainly not a consensus view, so I don't expect many to agree. Hence why I called it the most underrated tool in all of finance. And when I say underrated, I mean in practice, understanding that everyone with an undergraduate degree was probably taught about DCFs at some point. Most dismiss it as a theoretical construct, but if one never properly learns how to use a hammer one might conclude that hammers are not effective in driving nails or that nails simply cannot be driven.
“All models are wrong, but some are useful”
(1) "It is certainly not a consensus view, so I don't expect many to agree."
I think there may be a little bit of a misunderstanding. In your section on DCFs, you say how stock prices are simply math problems. This is definitionally false. No model output, whether it is a DCF or comps or whatever, has anything to do with what the price of a stock is. The price of a stock is solely determined by what it transacts at.
(2) "Most dismiss [DCFs] as a theoretical construct, but if one never properly learns how to use a hammer one might conclude that hammers are not effective in driving nails or that nails simply cannot be driven."
I am not saying that DCFs, in principle, should be dismissed as a theoretical construct. This is a valid way to understand a business.
With that being said, the standard way of doing a DCF is nonsense. All the (edit: information) that one actually cares about gets manually inputted by the modeler. That means you aren't really modeling anything. You are just basically a calculator. A truly useful model would be where the modeler places the target in, and the model itself will output all the information.
For example, instead of having the modeler put in the growth rate for the number of stores Apple opens next year, the model itself should spit out that number. Instead of the modeler manually putting in the price curves for crude oil, the model should be able to organically develop the price of oil for each time period without modeler input.
Fundamentally, the standard approach to DCFs is for every relevant operating detail to be manually put into the model by the modeler, instead of the operating details to emerge from the model itself. Assuming the details that you want to forecast is a nonsense way of doing analysis. It is what causes so many of the "bUt i CoULd nEvEr hAvE PRediCtEd tHaT" bits even though such an event/consequence is entirely predictable.
(3) "All models are wrong, but some are useful”
I couldn't agree more. A standard DCF model is of little utility in determining stock prices. Stock prices are not "simply math problems" in the way you claim they are
“All the inputs that one actually cares about gets manually inputted by the modeler. That means you aren't really modeling anything.”
Yep, you lost me on that
Did you read and understand the examples I provided where I made that idea far more tangible? I ask because the "All the inputs" part should be "All the information" but the examples I provided I hoped would help make clear that what I am talking about is how information is generated (ie - whether it is put in by the modeler or emerges organically from the model itself)
What emerges from the model organically is a stock price based on varying scenarios for growth, margins, reinvestment, and risk. That is the entire point of the exercise – we’re not trying to model Apple stores or the futures curve. Our goal is to generate risk-adjusted performance by buying and selling stocks, not forecast things for the sake of forecasting.
How those variables come together to form the output of the model, a stock price, is a function of arithmetic. I don’t disagree that fundamentally the price of literally anything is what people are willing to pay for it, but the DCF is a tool for understanding the price people are willing to pay – it is obviously not a gating prerequisite for market pricing (and I don’t think anyone else is reading my comments as that is what I am saying). The strength of the DCF is you can form the output by constructing various scenarios for those key variables and seeing what they imply for stock prices, or you can start with the stock price and sensitize the variables to figure out what is implied at the current market price (i.e. the market’s expectations). If we don’t agree that stock prices are the collective expectations of all market participants for a business’s future state then we probably disagree on how the market works.
Just because most investors aren’t each individually using DCFs to make buy and sell decisions does not mean that the DCF is “of little utility in determining stock prices”. Again, understanding that my views are not consensus views, I believe the market is an emergent system. Consider the old example of people at the fair guessing a cow’s weight. Most people don’t know much about cows; they’re not veterinarians or farmers etc. Some make their guesses based on how the cow looks, others guess based on something they do know more about and draw inferences from these comparisons (i.e. my car weighs X lbs so the cow must weight Y). Some guess seemingly randomly, unaware of the experiences/knowledge that may underly their guesses. Individually, the range of guesses will be very wide, but collectively it is very hard to beat the crowd’s average guess, because the crowd’s guess collectively captures little bits of truth about the cow’s size, shape, and weight that get to the true intrinsic weight of the cow. My view is that the price one is willing to pay for a business also has an intrinsic value based on the cash flows and returns one expects to earn over the holding period. But I also believe the future is nondeterministic, which means we need to model a number of different scenarios for the business’s future state and assess risk-reward probabilistically. The DCF is the only tool I know of that is effective in doing just that.
It's funny, no one argues on the math that underlies a bond’s pricing, but for whatever reason the same logic doesn’t apply to stock prices. But when you boil it down to the simplest level, you put money into an asset with hopes to extract more value than you put in. Everything comes back to cash flows.
I also will say that I cannot definitively prove my views to be true. All I can tell you is that based on my experience, the DCF is one of the most powerful tools I have in my toolbox for generating risk-adjusted performance after 10Y of being in the seat and having started off largely dismissing the DCF when I had less experience. I don’t know your background, but you wrote your comments authoritatively. I’d ask you what you are basing your conviction on in some of the statements you made?
That flawed logic is why active LO continues to underperform market. No wonder smart money continues to flood to ETFs and pods as a result.
passive investing has grown too large, we are now in a feedback loop which will have a dramatic reversal at some point.
thanks, thought-provoking read. i'd be more curious to hear how your career's gone- are you still at the same fund or did you move, do you have line of sight to managing capital directly, etc
I’ve changed funds once in the 10Y. I feel incredibly fortunate to have done so by choice as I would say a lot of the people I have met in the industry have had to change seats involuntarily at some point in their careers. I started out on the insurance side and now at a traditional asset manager. When I left I would say things were trending very positively for me, but I had a greener-grass question in my head that had to be answered, in addition to the fact that my current seat came with materially better economics.
On the insurance side, you have one client – the insurance company. This gives you permanent capital, and absent a CIO change where someone comes in wanting to take everything passive, it is a pretty stable seat. Your flows in and out are based primarily on how the company is doing, which is normally pretty predictable outside of major CAT events or unexpected market share losses.
On the asset manager side, I don’t feel that same sense of security. You have a tough stretch and you start seeing the redemptions. Even when you have a good stretch, if you don’t have the sales and marketing heft to get the performance in front of people you’re not going to grow. This is something I could have put in the original post, but sales and marketing support is absolutely critical to a fund’s longevity. On the most extreme side, Cathie Wood was down -70% the first half of 2022 and still got positive flows. Regardless of what you think about her, she put the narrative through the spin cycle – heavy load, and that ability to craft the right narrative at the right time might be a more important skill to the firm’s economics than picking the right stocks. When things are good, you gotta raise assets because you are definitely going to lose assets during the tough stretches.
I’ve produced good performance since coming over and feel that I have my process pretty dialed in at this point to something that is repeatable (though always iterating). I think that has been recognized, and I have pitched a product idea where we have shelf space, but my biggest concern is having the horse shot out from under me given struggles on products that I have little control over but that drive important economics for the firm. Committed to trying to make things work here for the long-term, but we need to fix some things is the short answer. If we can weather our current storm, I’m pretty optimistic to how things will evolve, but we’re facing some adversity at present.
Good post, I liked the part on taxes as that's 100% the reason a lot of flow-based funds are still around. Whereas most institutional LPs have started to realize that pods like Citadel and Millennium are doing actual hedging and pro-LPs vs. other archetypes which mostly long beta and pray.
I get what you're saying on "underrated" but I still disagree on the DCF argument. Every good PM will tell you that in their years of investing, no one ever made money from two pitch types 1. a SOTP argument without a clear catalyst from an activist to break up the company GE style and 2. a DCF model where the inputs such as WACC and cost of equity are completely made up and if you pull up, you'll realize a DCF is inherently circular.
You can also completely automate DCFs using AI, a startup recently did this, and found zero correlation or backtesting power between "good DCF" stocks and "bad DCF" stocks across the 8,000 or so public universe. In fact, the correlation was slightly negative e.g. -25%.
That's because in this day and age, this is a backwards looking tool that has no predictive power with respect to how the company's market share and fundamentals would perform relative to the numerous tariffs, bills, AI-driven changes, new competitors, and other temporary to permanent shifts from major drivers in that company's sub-industry. Take a look at Chegg for example - would a DCF prior to AI disruption help you make a 3-5 year investment in that name? Or share your Tesla DCF model. There's a huge bear and bull debate over robotaxis and it's a series of logic trees with assumptions ^ 3 as an input that if you were inclined to model, would spit out DCFs ranging from $10 a share to $5000 a share.
The better exercise here is the tried and true approach of figuring out what's priced in, what's incremental, and what moves buyers and sellers. At the end of the day, that's what moves stocks aka buyers and sellers. What is implied in expectations and do you have a non-consensus view. TLDR: if DCFs had any predictive power, the best investors in the world from both quant and long/short funds (Rentech and Citadel) would be employing it even as a side function. During my time at Citadel I didn't touch a single DCF and neither did any other analyst team.
The only times DCFs were useful were in PE when you had majority control and needed the cash flows for obvious reasons such as debt waterfall modeling and in that case you can make a pitch that you run a DCF if you KNOW that a sponsor is about to take-private a company and you want to quantify upside since you know they're building a LBO model. But even in that case, you wouldn't even bother if you 100% knew a buyout was happening just buy the stock regardless and also you would actually run a LBO model to imitate what the sponsor is doing.
Completely agree with your comment that what is important is “…figuring out what's priced in, what's incremental, and what moves buyers and sellers. At the end of the day, that's what moves stocks aka buyers and sellers. What is implied in expectations and do you have a non-consensus view.” I think I’ve written almost exactly those words in above comments to others, but my view is that the DCF is an underrated and powerful tool for doing just that. The DCF has no predictive power in and of itself.
RE: startup, as a starting point, I can tell you that it would be impossible to automate a DCF, at least in the way that I have talked about utilizing it, and doing so would have zero utility as that really isn’t the point of the tool. So the fact that an automated DCF didn’t produce a good backtest is not really surprising, but it is a little like trying to cut a 2x4 with a butter knife and then saying knives don’t work. Interested if you have a link on this though, would like to read for myself and don’t want to just dismiss it without knowledge of what you are referencing.
RE: TSLA, I agree that it is incredibly hard to model and you have to believe in a lot of things going right at this price that are extremely difficult to forecast individually, let alone in aggregate. But that is the point, and we do not have to invest in every stock in the investable universe. For a disciplined investor, it is ok to put some things in the too hard bucket. Conversely, if you do buy a stock like that, you’re making the forecasts whether you’re aware of them or not.
RE: Citadel, they’re playing a different game than I’m playing, and I’m fully aware that I am never going to have access to as much information or processing power as they have. I’m never going to be able to dunk over Shaquille O’Neil, so I’m just not going to play him in basketball because I like playing games I can win. I have a lot of respect for the Citadel model but there are many ways to make money in the markets. The fact that you didn’t come across Citadel analysts using DCFs does not really matter in my model of the market as an emergent system, discussed in above comments.
I’d offer another perspective on this topic. There was/is an old CS piece that cites a survey of analysts and showed that 99% of them use some sort of multiple for valuation and less than 13% used any variation of DCF. Regardless of the exact numbers, I think we could agree that it is true that most on the buyside do not use DCFs. Thus, what is the more compelling observation? That Citadel analysts do not use DCFs and that Citadel has generated good performance, or that most in the industry do not use DCFs and most in this industry underperform net of fees?
By extension of that, while Citadel is incredibly successful at the aggregate level, we know that the average tenure for an analyst at the pod level is maybe 1Y to 2Y? The analyst/pod level is characterized by a high rate of failure. So perhaps the success of Citadel is what they are doing from the top down and less attributable to the fact that analyst teams are not using DCFs.
Would you mind commenting on what models you were using at Citadel instead of DCF? Perhaps an example using hypothetical stock ABC?
Interestingly, I didn’t expect this post to turn into a roasting of DCFs, but to quote Seinfeld “you know the very fact that you oppose this makes me think I’m onto something”
I agree with basically everything you've said here. Also ~10y exp (5 buy side, 5 sell side).
Some people think DCFs require too many assumptions and they are too sensitive to those assumptions, but I don't think they've really thought about all of the choices you have when it comes to the alternative: multiples.
Great post, thank you for sharing! Curious, if sell side consensus isn't what you look at to form a view on consensus, what is the best way you've found to get an idea of what buyside consensus (i.e. the consensus that matters) actually is to verify/quantify whether you have a variant view? Thanks!
Sell side consensus is certainly a good starting point, but not always what is priced into the stock, i.e. the buyside consensus. Best way I know of to get a handle on buyside consensus is to reverse engineer the market price through a DCF, but you can also think through the revisions before they happen and consider what that implies about the real multiple the stock is trading at. I’ve defended the DCF quite a bit in this thread, but the best approach is typically multi-faceted.
The best example of when sell side /= buyside is on deeper cyclicals. This is why cyclicals look expensive near the bottom and cheap near the top. The street is slow to update estimates for cycle dynamics as they typically need permission from management via updated guidance to make larger revisions. But the buyside starts pricing in an inflection well before it is reflected in sell side consensus, and thus, multiples of sell side consensus will be counterintuitive for this type of stock where there are larger swings in earnings and growth.
Once you’ve established a reasonable approximation of what the market believes, you can compare that to what you believe and start to identify differences. But that’s not enough, you also have to try and figure out why the market believes what it believes, why your views could be better, and if/how the market’s expectations are likely to change. If you cannot identify a mistake in the market’s pricing, that would be a signal to go back and do more work or re-examine one’s assumptions. Once you’ve done all that work, you’ll know whether you have a variant view or not that is worth betting on.
Thank you, this is really helpful! A few things about reverse DCFs I've always been confused about though practically:
Just want to check in on how you think about practically implementing reverse DCFs as I've never built one out personally because of the above, but I might also just be thinking about how to build them completely wrong. Thanks!
1) Why CFA and MBA? What's the point and what's the value add? Is it just because it looks good on the company website and in marketing brochures?
Is it because at the end of the day, the business is mostly marketing and optics ?
Is there actually a value add to CFA and MBA from the education and knowledge side? Because I don't think there is.
I have a CFA and it's been useless. The only thing I picked up / formalized was the concept of human capital when doing financial planning / personal finance. I do think it shows that you're somewhat smart, and can learn a lot, and have learned a lot about everything at one point. So there's a base level of knowledge. But most people forget everything after a week
2) "But you have to have a handle on 1.) the math that underlies multiples; and 2.) if you are comparing a current multiple to a historical range, the numbers that were underlying those multiples earlier in the company’s history. You cannot simply look at a chart and say a stock is cheap or expensive because it is north or south of some line drawn on a chart. A multiple is a direct function of expectations for future ROIC, growth, reinvestment, and risk."
^ I think agree with what you're saying about multiples vs. DCFs. When I first started, I had a handy "multiple" calculator that spit out a multiples based on rev growth, ROIC, ROE, margins, taxes, dividend, discount rate, etc. Multiples are basically a short hand for DCFs.
After you see a lot of companies and industries and models, you understand why they get the multiples they do, whether it's P/E, EV/revs, EV/gross profit, P/B, etc.
A more simple, in-between compromise is modeling out a company 3-5 years out and seeing what multiple it trades.
I agree the important thing is understanding the math and drivers behind the numbers at a given time. Obviously there are billions of variables, so it's garbage in and garbage out a large extent, but it helps keep you grounded in some sort of reality.
CFA because it’s accessible to really anyone for a low financial cost. It does give you a baseline knowledge of financial markets but it doesn’t make you a better investor or even really ready to do the job. But at the end of the day I largely agree that it is mostly marketing and optics, and it gives you an edge only if your interviewer also has it (which most in the industry do on the LO side). I don’t have any nostalgia for going through the program, it was a major time suck and pain in the ass, but I can honestly say it directly contributed to me getting my foot in the door because everyone I ended up working with in my first seat had it and emphasized it. It is what it is on the CFA.
I think we are largely on the same page. I’ve done enough DCFs at this point where I can pretty quickly get a feel for what are likely the key sensitivities that are going to drive the multiple just based on the type of business it is.
I’m associate with 2 YoE at a T1 LO and I’m not required to pass CFA.
Would you still try it? What are the advantage to have it in my situation?
I feel like time is better spent by learning my sector at the moment
Harum unde quos error aut. Consequatur earum rerum voluptatibus ex aut omnis.
Sunt non temporibus laboriosam qui consequatur ut atque. Mollitia qui recusandae neque laborum ut accusantium sint sed. Ut quibusdam quia qui cum similique qui cumque.
Omnis maiores hic sed illo. Nobis in pariatur molestias sit labore tenetur dolores. Voluptatem porro aliquam nesciunt quam assumenda accusamus est doloremque. Illo error ut alias qui maxime.
Dolores quo quam harum. Consectetur minima eum quas deserunt molestias. Possimus necessitatibus maiores molestiae dolor vel distinctio suscipit sed.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...
Rerum delectus omnis perferendis sequi facilis nemo. Dolorum quo ullam aperiam asperiores. Quam exercitationem omnis non odit animi esse.
Non veniam quis sunt porro ut. Labore est sed et aut eos qui. Neque qui qui dolor repellendus neque. Architecto incidunt ut iusto tenetur ut quos necessitatibus et. Iusto enim dolore dolor.