Do you believe that in 20 years you will need to be an exceptionally quantitative person to work in this industry?
I just wrapped up my summer internship in equity research and after I completed and presented my final project, the head of the small firm I interned at pulled me aside to discuss my career aspirations and college major choices. He said if I want to work in the hedge fund industry, the most important thing is that I gain quantitative skills because "in 20 years, I'll be obsolete even in equity research, everything is moving towards quant." Is this accurate? Does everyone need to study math if you want to get into investing at a professional level?
20 years from now is impossible to predict now, given how fast artificial intelligence is moving. For what it's worth, though, what do you think he would predicted over the *last* 20 years? If you asked that guy back in 2002 what he predicted for the industry over the next 20 years, what do you think he would have said?
I bet the advice back in 2002 would have also been "everything is moving towards quant, you need to study math", too. And yet, here we are today.
Fundamental equity research will continue to exist for a long time because you will need humans to read and interpret financial statements (algos are not very good at this). Despite the fact that accounting is all numbers, it's about as far from objective science as history or English.
However, adding more quantitative skills (including math and programming) will never hurt you and can only give you optionality in the future.
My prediction -- investing will get more and more quantitative for the next 5 years or so, but after that, the artificial intelligence needed for investing will be so advanced that only experts will be able to maintain it, so it doesn't matter if you took basic programming and math classes or not -- it won't do you any good unless you have PhD level knowledge. I think there will be no value whatsoever in humans reading financial statements cause machines will do it better. It will be nearly impossible for any human to outperform the computers solely through financial statements (or even publicly available alternative data like credit card info).
However, there will still be a niche for humans in providing "market color" (which basically means networking for rumors and gossip that don't show up in official financial statements. Some people would call that insider trading, but it mostly stays on the legal side).
Humans will not be able to compete with the machines on exchange traded equities. But humans might still have a future in less liquid assets like corporate bonds or mortgages or physical real estate, cause those rely on information that is not easily searchable (and therefore harder for machines to scrape).
Great take, completely agree. My forward thinking philosophy for my career is to be in a position of creation. Humans will either be creators, owners of IP, a business, or they will generally be obsolete. The rapid change we’re seeing across job functions in the broad corporate world will continue to increase.
It’s a dual edge sword, a sort of paradox. You need to work at the corporate level to afford a decent living and fund your future entrepreneurial ventures; but the longer you’re in the corporate world rapidly evolving and learning new programs, ways of working with the algos, etc., the tighter the squeeze becomes between you and the machine. Very interesting times
Most people correctly understand that there is a substantial qualitative component to investment judgment that can’t be coded away because the inputs are not discrete or structured and can’t be made so (without humans at some pt in the process). For example, so many data vendors sell sentiment data and I think it’s mostly crap - because volume of tweets or good/bad words doesn’t mean the same thing every time.
But I think what people miss is that teaching learning still has to have its basis in a human idea. I agree structured and uniform and commoditized data like mkt data (mostly done), financial figures (done to the extent it can be, what’s left is all the changes and md&a and resegmentations and messiness). But just like primary research, the real durable alpha is and will remain based on creativity and divergent approach. There are so many free, public data sets over the years I’ve found and some been able to use in a way that people just didn’t think of and that’s where all the money is made. The PHD quant genius can’t machine learn stuff like - he doesn’t know the business well enough to have that kind of divergent view on how to use, say cell phone pings (commoditized) tied to cruise ship itineraries and incorporating rental car stats in cruise ports on departure days vs. non-departure days (this is a fake example cause I don’t want to give a real one)
The fundamental guys will never win any game that can be won purely with speed and data and efficiency. But on a medium/LT time horizon - there will continue to be alpha for humans and imo where fundamental/quant is mixed (this is expanding) the quant compliments and scales the fundamental insights. The kind of real variant views that this quant work enables and expands is the same type of variant views as when we hear about Julian buying and deconstructing a car or Einhorn visiting a shrimp boat. The stuff you’re talking about is different - it’s like when people talk about “we used to just be the only one who asked for a 10k to be mailed” or Tudor when he was the only one with a guy to make charts or old school Rothschild who had carrier pigeons to know who won a war first. Everyone knows that who wins a war matters for trading sovereigns - even back then. It was the scale and efficiency that was the edge. There are too many scenarios, in fundamental investing where the thing that matters is too one-off or unstructured to be coded into a uniform project cause it won’t scale and hence the roi is low. These opportunities will continue to exist.
while your comment did not go the extreme, I sometimes see people say that one day everything in markets will be quant and ML. That’s akin to me saying that one day we can model partical physics so accurately that we can run a simulation forward and see the future. First - the technology is further off than you think. Second - more importantly - even if we could model every atom of an explosion or car collision or chemical reaction perfectly, we still couldn’t project forward the world becuase the subjective, thus far immeasurable/modelable/structurable human element - that is ultimately interacting with all those explosions and car crashes
This is a bit of a ramble. If I could go back in time to college today and restart my career in 6 years - I’d choose fundamental side again
are you in equities or FICC?
Laborum ipsa impedit in cum aperiam iure eligendi quis. Placeat eum magni in ad commodi. Accusamus voluptatum perferendis sunt.
Sed porro at et exercitationem dolore. Libero aut maxime delectus est repudiandae ducimus nihil. Magni soluta expedita et rem harum ex magni. Et voluptates soluta qui sit.
Ipsa ea sed possimus cumque. Natus tempora doloremque eos quasi illo quaerat est. Excepturi alias ipsa voluptatem est. Corporis dolor hic soluta magnam ea. Officiis eius magni sequi ad.
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...
Debitis ut illum voluptatem fugiat quae quae et. Nulla officia voluptas veritatis rem. Qui dolore consequatur eos dolorem soluta.
Odit maiores et repellendus hic recusandae in. In recusandae reiciendis et porro nesciunt explicabo. Quia minus amet enim est aut dolorem.
A ipsum quia consequuntur quod. Ea esse et aperiam eaque. Dolore ab amet amet. Atque quae voluptas corrupti ipsa exercitationem dolor id. A molestiae nostrum corrupti rerum. Nihil dolores suscipit quas aut.