The most inspiring talk at today’s Buy-Side Technology Summit was one hedge fund manager’s passionate appeal to get rid of all IT staff, a comment that generally did not go well with the predominantly tech-focused crowd. Simon Hazlitt, cofounder of Majedie Asset Management argued that there’s no reason for funds to waste money employing technologists; there’s no reason to maintain cumbersome infrastructure and proprietary systems when it’s so much cheaper and often more efficient to go to the cloud. In Simon’s paraphrased words, IT integrated with business doesn’t work, the majority of project costs are attributed to monitoring project costs, and really, all projects are “rubbish.”
Outsource All The Things
For those unawares, cloud computing effectively refers to outsourcing computer resources to a third party. For example, rather than storing files on your computer, you could store files on the cloud (via a service like DropBox, or Apple’s just announced iCloud), and then those files become accessible from anywhere with an Internet connection. Even if your computer breaks, the files will still be available. The cloud is (almost) always available, and easily scalable. For more details, see Wikipedia.
Surprisingly, the follow-up panel, which consisted of professionals from across the industry, rather than fighting back in defense of high tech in financial services, sort of agreed: albeit much more slowly at bigger firms, the move to the cloud is happening. Why spend money managing an internal messaging infrastructure when you can move to gmail, for example? (NYU recently migrated all of their collaborative systems, including email to Google.) Following the argument to a natural extreme, why build a proprietary trading system, when you can use a vendor platform running on the cloud?
Though Hazlitt made an interesting point on delocalization, that cloud computing is a natural extension, of the suburbanization, created by the proliferation, of automobiles from the 1960s… in my view, “cloud computing” here can generally be synonymous with “outsourcing.” My outsource argument, which Hazlitt would likely agree with, is based on the argument of division of labor and economic interdependence. In most cases, I’d bet there’s no real need for financial services to employ a full IT staff (as a technologist in financial services, it’s ok that I say this, right?) – financial services could focus on investment decisions, and engineers at vendor firms could specialize in building domain-specific applications.
Specialize
Specialization in this case would lead to economies of scale, lower costs, and I think quite likely more innovation. Yes, there are arguments against this idea, such as the need for very detailed firm-specific application customizations, for people to translate business requirements to vendors, to configure things, or the occasional need for proprietary technological logic (such as statistical arbitrage algorithm implementations); however, you wouldn’t expect a dentist to build his own dental equipment… Anyways, I digress…
Other Tidbits from the Buy-Side Technology Summit
- Forecasts in spending see more allocations in fund administration and reporting, consolidation across applications, and in the mobile space; in contrast, it was noted that forecasts see cutting back on “Park Avenue wages for people processing trades.”
- Josh Levy with Tactical Asset Management described execution management as deciding “which cabin to sleep in on the Titanic” alluding to the fact that in a sinking portfolio, chasing pennies makes little difference; there were, of course, valid responses to this.
- There were some heated discussions around high frequency trading and how quote stuffing may or may not be a good thing or inevitable for efficient markets.
- Every panel emphasized the importance of Transaction Cost Analysis (TCA) and Venue Reporting, noting that it’s critical for traders to understand how algorithms and dark pools behave or perform.
2 comments
The reason for in-house IT projects is…
… if you were an IT manager, would you advise senior manager to cut your budget and outsource your job, and/or jobs of your subordinates? A smart IT manager is very likely to feed budget authorities with well-meaning lies that inflate the importance of his division. Incidentally he might be sincere while doing this – personally believe in the BS of his own making.
In my experience most typical argument against outsourcing of in-house IT projects are based on misguided security assumptions, i.e. lay people are conditioned to believe that data stored in the cloud is less secure, so budget authorities are likely to take this BS about security on faith, without asking questions.
The reality is that argument of ‘lesser security’ is conceptually bogus: one can only apply ‘less’ or ‘more’ operations to quantifiable qualities. To make an intelligent claim that one system is more secure than another, first one has to measure security of both solutions. To my knowledge, no such metric exists – there is no even Unit of Security, as of yet, to make such measurement possible.
One of the many reasons security can’t be measured is because it’s a very vague concept. People who are less clueless, would probably avoid talking about ‘security’ in general, but talk instead about tradeoffs b/n specific risk factors. Yes, if you don’t store your data with Google, the risk of your data being compromised because of incompetence of Google employees is zero. Yet this leads to a higher risk of your data being compromised because of incompetence of your own employees. No mater what you do, you are never risk-free, most of the times you run away from the risks you are aware of, right into the risks of which you are blissfully unaware.
Hey Misha –
Thanks for checking out my experiment in social media here!
I completely agree with you, and your point on security is something that was addressed during the conference. The argument was, as you noted, that in any industry where the security of information is critical to business, you have to put your trust in someone, and that someone can be internal technologists, or it can be a vendor product. There are different risks associated on both ends, but the cost of doing business in the cloud can be significantly cheaper for a given perceived level (“feeling”) of security.
Comments are closed.