F*** infra modelling
The title says it all. Spent the last 4yrs in Infra PE and can't stand how much we get bogged down in the details, knowing that we always end up miles away from the truth. Is it the same across the other verticals (tech, healthcare, industrials)? Any industry building simpler (straighter to the point) models?
as as fellow 3.5 year infra modeler I feel this all too well.
Is it ever satisfying?
Could be worse…you could work on the generate equity investment team and now everyone knows those guys can’t model…or do anything well in reality
Will be interesting to see where they land…
But hey hold on now, they are rebuilding the world (together) and keeping us below +2°C!
I’m pretty sure they had PJT re-do a bunch of their files when they were fishing for capital. Godspeed
this is so unbelievably not surprising if it is true…
My piece of advice to all students when I was in banking, was always to avoid infra related sectors as you sign up to a life of pain. You can’t leave banking hoping for a better work life balance because infra PE is even more miserable. You’re gonna spend your life working like a dog on 30y models even as a director
Let’s run the 340th tornado chart of the week and see how our 12.7863% IRR changes
Do you have any views on which verticals spend the less time on modelling? (Still talking investment related)
Model granularity can really depend on a number of things like complexity of capital structure, platform vs. asset, etc. As a generalization I’ve found that power & renewables (tax equity) and utilities (rate base modelling) to be the most painful and digital infra to be more palateable.
Unpopular opinion but I actually enjoy the detail. FWIW I was a generalist and healthcare was just as granular, so the grass isn't much greener. At least my experience.
Thanks for sharing, super helpfull! Would have imagined that health care is in a similar position. Are you aware of any verticals with lower modelling intensity? Tech, consumer, industrial?
Retail & leisure and consumer are on the lighter end, tech varies a lot (SaaS is clean and structured, hardware/platform gets heavier), and industrials sit somewhere in between the mentioned and Infra
Software PE guy here... you guys are building models?!?
Nice one haha
Have worked in infra PE at a mega fund for the past few years and I gotta say, this may be the most brutal form of investing from a lifestyle perspective. In theory, all of the businesses are very simple - contracted or visible revenue streams for multiple years. You are usually buying either a yield business or a platform business where you are going to leg in capex over the hold.
What makes it fucking awful is the models are gargantuan and the partners want to sensitize every goddamn variable and case you could imagine. “Can we have a downside case where x asset takes 2 months longer to fully come online, and how about we run another scenario where we factor in a 50bps pricing hit 5 years in the future at this site”. It’s absolutely retarded the amount of granularity that we model just to be wrong. And on top of this you have to uncover every rock and pebble during diligence and analyze every kpi and operating performance metric a bajillion different ways.
Does infra PE make you a more technical modeler? Yes, probably. Does infra pe make you a better modeler and investor overall? Probably not. It is so details driven while unreality there are usually 2-3 real drivers that are going decide the outcome. Would avoid this shit like the plague. I have been in a ripper of a live deal over the past 3 months and am pushing 100 hour weeks of HARD fucking work. So has everyone else on the deal team if that is any indication of if things get better.
The granularity is no joke. A lot of those models especially like regulated utilities start blurring the lines between a spreadsheet and software and you need macros to even interact with it etc. Hang in
Absolutely spot on! Glad (and sorry...) we are on the same boat. Are you thinking of transitioning to another industry?
Yes, 100%. This shit is honestly unbearable. As others have mentioned, the value prop for MF investing is just not there anymore. If anything, it’s even worse for infra investing at these shops because hold periods are longer so realization of carry is worse. Also our return hurdle is still 2.0x to trigger the GP waterfall so it can be like 10-15 years for a fund to payout. It’s fucking stupid to bet your career on that. Also the chance you actually make it to a partner seat is so slim.
On top of all that, the work generally just sucks. Infra businesses are very boring. After looking at the nth renewables or data center platform, all of these businesses are the same. Buy for shitty multiple, optimistically think you can deploy a bunch of capex over the hold accretive, and then sell for a similarly shitty high multiple. My fund is basically just levered long data centers and renewables tied to data centers at this point.
Do you think moving to a pension / SWF or infra credit would be better from a lifestyle perspective?
Not really, but it depends. The Canadian pensions (CPP, OMERS, PSP, Teachers) and SWF (GIC, ADIA, AussieSuper) that are credibly underwriting directs are still sweaty. They are not as bad as GIP, BX, EQT, KKR, etc., but they’re not lifestyle seats like many of their counterpart groups in the same org. I don’t have a huge level of visibility into credit but I know GIP credit (team investing out of CAPS and Spectrum) as well as BX infra credit (BGREEN) have terrible hours and culture and are a revolving door. These guys are in the market for seats like every 6 months trying to plug the hole.
Edit: One thing I’d add as well, and not dissimilar to their MF PE counterparts, but the value prop of going to a MF beyond a 2-3 year associate stint is just not there anymore. The industry is big enough and there is enough capital flowing into these funds that there is probably decent upward mobility at most funds through to a VP/Principal seat. However the math just doesn’t math on getting any meaningful carry that justifies the sacrifices (and frankly luck) that it takes to get beyond this point. If I could go back and do it again I would’ve lateraled to a smaller fund and shot for the upside.
Hey man I am about to start working in Infra PE any chance I can PM you
Nearly all models are wrong but some are useful is what I’ve learned
AI will improve tremendously the workload on these infra models I believe in the next 2-3 years ..
Think the nature of infra (finding "edge" in assets that are highly contracted and regulated) just means the time will be filled running more sensitivities, more analysis... agree AI might physically make the model but that just gives you "leverage" to spend your time looking at a nearly infinite number of analyses.
Agree with the messaging. Yes, AI could alleviate part of the grunt work but it also opens the door for more random sensitivities. Can't wait to run monte Carlo simulation on all 5,000 inputs of Assumption tab number 1
Good meme.
Infra is for elite only, no other sector understands cash flow and non recourse finance as well as for this group. Rest of you are carpet sellers
I know you’re trolling - but you’re legit right actually lol
Lol i feel you too well,
But i kinda enjoy the granularity (party because i'm a bit OCD about my models). Yes, the VPs/Ds/Ps will absolutely bomb you with the most idiotic scenario possible but if you've done the model correctly from the start, your model should robust enough to handle every little BS scenario your superiors throw at you. Even if the IRR would only move 10bps.
The key is basically owning and mastering the Project/Model itself that you could give a estimation of outcome without even touching the model.
VPs wanted to see what if the tariff increase scenario decreases by 50bps every 2 years? Should be around -30bps decrease of IRR because the discount rate compounded for 50 years.
Ds wanted to see how FX rate changes impacts IRR? Most currency pairing has stable "spread"/"depreciation" over the years, apply those number and your IRR might only be 10-20bps off the full calculation within the model
Ps wanted to finance 90% using debt instead of 70%? The IRR difference should be the cost of debt value delta between 70% and 90%
Most of the analysis outcome is quite easy to predict if you already in the space for some time and having those mental math down will considerably save your time and in some cases your bosses will rather take your word (and you get recognition) rather than asking you to actually run the model.
Agree to some extent on the flexibility though disagree on "mental math". At least in the different teams I've been, no one ever got satisfied with a rough estimation. Every single scenario has to be run so even if your mental maths are correct, they would ask you to run everything no matter what
Deserunt velit ipsam tempore tempore. Consequatur exercitationem expedita voluptatum quidem nesciunt est. Veritatis ut aut dolore eius possimus iusto amet. Voluptas quidem deleniti ut voluptatem odit.
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...
Velit sunt non provident aut unde. Adipisci culpa et voluptas nihil et officiis explicabo. Est excepturi nihil necessitatibus qui repudiandae dolorem beatae. Vel blanditiis dolor assumenda et suscipit.
A sapiente consequatur explicabo. Ut ipsam facere id voluptatem. Molestiae pariatur et doloribus unde numquam corporis quo. Tempore a corrupti laborum eum occaecati.