UBS and the Chip

Date: September 28, 2011

When we first faced the question of switching to Cloud technology from traditional technology, the biggest question was around the business model. We were keen from the outset to change the model and adapt it to the technology. We ended up deciding to offer all the core functionality of StatPro Revolution to anyone who bought just one portfolio. Why was that?

A long time ago (approximately 20 years) when I was working in Switzerland, a client of ours called Nicholas Weiss who worked at UBS told me an interesting story. Apparently, back in the 1970’s when banks had to rely on the mainframe computers supplied by IBM and a few others, UBS had a real problem processing data fast enough to satisfy their needs. Their burgeoning IT department had tried all means possible to secure faster processing, but failed. So they did some analysis and worked out that it would be possible for them to build their own computer chip to their own specifications that would run fast enough for their needs. So they put together a business proposal and presented it to the UBS board to consider. This was taken very seriously and there was much debate, but in the end, the board decided that they should not get into the business of manufacturing chips.

Nicholas and I obviously found this very funny in the oh so modern 90’s and laughed a lot about it, but it is an interesting story for any period if you think about it. At the time, I was trying to sell Nicholas a standardized off-the-shelf database solution for UBS’s mutual fund data, but the competition was the in-house development team who thought they should build it. Apparently the project was code-named Project 5000, but had been going on for so long that the internal joke was that the name was the date of expected delivery! Nicholas knew that we would provide something useful for him very quickly and very cheaply in comparison to the internal solution, but it was a big battle internally.

These days, there is very little internal development of major applications as companies have learnt their lesson. The technology goes out of date quickly, internal development is often less well documented and when key people leave the project often freezes. Worse, the investment can only be amortized across one client. Of course when there is nothing on the market, it can be very frustrating, but in this case it is often better to encourage a start up to make it than build it internally.

So what is the modern dilemma? Well obviously it is the decision so many boards must be facing right now between continuing to invest very large sums in traditional Client Server based applications in their own IT centres with all the associated complexities or to trust to the new technology of the Cloud. The short term issue is that Cloud based applications are relatively new and don’t cover every type of key application yet. What to do during all this change if you can’t find the Cloud based solution that meets your current requirement? Buy new applications in old technology which you know will become embedded in your business for the next 10 years and be increasingly expensive and inefficient or sweat the old system for another year or two until a cloud-based application appears?

In my opinion the Cloud is such a massive leap forward in productivity that nearly every sort of application will switch to it over the next few years. Once you run an application built for the Cloud, that application is far easier to maintain and keep up to date, so you may never need to switch again. So when you are asked to make a significant investment in a new application, my view is to ask if the proposed new solution is in the Cloud and if it isn’t consider keeping your existing product a little longer to see what may come. Otherwise you may regret the decision in one or two year’s time.

I am sure this is the same view the UBS board had back in the seventies about that chip.

To learn more about 13f-2 watch our webinar replay Part 1: Unpacking the SEC's New Disclosure Rules for Shareholders
Join us for Part 2: Operationalizing the SEC's New Disclosure Rules, for Shareholders on December 12.