Tag Archives: hedge fund operational issues

Data Warehouse Implementation | Hedge Fund IT Solutions

Deloitte & Netik Conference – Asset Management: Managing and scaling your reference, market and portfolio data

This article is part of our guide to Hedge Fund Business & Technical Issues.

On October 29, 2009 I attended an event at San Francisco’s Omni Hotel put on by Deloitte and Netik.  The event featured a panel discussion with the following panelists:

  • Brian Lott – Executive Vice President, Operations, Netik
  • Michael Smith – Senior Manager, Deloitte & Touche LLP
  • Joseph Clark – Market Data Manager, MSCI Barra
  • James Wu – Product Manager, MSCI Barra
  • Moderated by Ray Iler – Northwest Pacific Hedge Fund Leader, Deloitte & Touche LLP

The discussion was quite interesting and was most appropriate for hedge fund managers who have a need for transparency into their positions and who also need the ability to quickly manipulate large amounts of data.  Coming into this discussion I was not aware of any of the issues involved with data warehouse implementation.  I am sure that much of this discussion went over my head, but I did the best I could to summarize the points made by the various panelists.  As always, any errors or confusion found below is likely attributable to me solely.  [HFLB note: I have not checked with any of the panelists regarding my summary of their comments.]

****

Data Warehouse Hedge Fund Discussion

[Moderator questions in italics.]

Ray Iler started the panel discussion by stating that investors and regulators want more information, generally for risk management purposes.

Michael Smith – pending investment adviser registration will mean that managers need to make sure their data systems are ready to handle SEC requests for information, SEC audits and most importantly investor requests for data.  Data warehouses help asset managers to reconcile their various data inputs (especially with regard to multi-custodial relationships).

Brian Lott – regulation is going to drive what “transparency” means.  Old systems were not designed for the way that business is now done – data warehouses now allow managers to bring all information into a central repository and helps the manager to have a holistic view of their portfolios.  These data warehouses can also help ease the regulatory burden by speeding up the time it takes to search and categorize data.

James Wu – data warehouses give managers the ability to see and analyze portfolio and asset level risk characteristics.  These solutions allow managers to run robust reports and ad hoc queries fast.

Ray Iler asked how data warehouses can help managers such as fund of funds or hybrid funds.  [Or something to that effect.]

Brian Lott – legacy apps have not evolved quickly over time.  Data warehousing goes beyond position level down to the data below.  Prior to data warehouses there were systems which were modified to try to function like current data warehouses.  The current data warehouse system is equivalent to a hub and spoke system where the data warehouse is the hub and the various data inputs (information or fees) are the spokes.  The advantage of this hub and spoke system is the ability for managers to centralize data and then be able to analyze it.

Michael Smith – it is one thing for a manager to understand the economics of an investment at the beginning of the deal, then then understanding what is going on with that deal on an ongoing basis becomes harder.  Data warehouses allow managers with a common platform to understand their deal on a continuous basis.  One application for managers is that they could understand how their portfolio composition would change based on an investment.  Such information helps a manager to understand the investment on a continual basis.

Joseph (Joe) Clark – another thing about clients without a data warehouse is that they have a hard time trying to find the data.  A data warehouse makes finding certain data much easier.

James Wu – one of the great things about a data warehouse is that allows you to have more spokes – legacy systems which have been modified (shoehorned) to try to meet current demands often are not able to seemlessly integrate new feeds.

Michael Smith – [made a statement about Microsoft sharepoint and other solutions which make technology easier.]  Managers moving into new asset classes (say a private equity fund moving into real estate investments) should think about how cash flows and reporting are going to change from current systems.  How will this change the formatting?

Joe Clark – data from different vendors and from groups in different jurisidictions do present an issue for some managers.

Ray Iler – that is good point – what about licensing issues for the feeds into the data warehouse?

Joe Clark – It is an interesting issue that centers on the question of who owns the underlying data.  Questions arise as to how many deriviations of the data are needed before the feed provider no longer “owns” it.  It will really depend on each individual relationship.

Brian Lott – it is not standardized from vendor to vendor and the issue needs to be negotiated at each data provider.

Ray Iler – what are the costs and ROI by putting a data warehouse in place?

Brian Lott – the question becomes, how do I pitch this to the board?  It is often hard to quantify the return but recent market events have shown why it is so important for managers to understand total counterparty risk.  For instance, after the Lehman collapse a client had to go through 72 different relationships with Lehman to understand what the total exposure was – this took 3 months to complete.  Generally not until a catastropic event do you see the value at the most basic level like this.

Ray Iler – although we understand that each project is unique, are there some common implementation costs?

Brian Lott – on the data warehouse implementation, for a medium buy-side manager, you are looking at 3 months depending on complexity.  Costs will be all over the board and will depend on the licenses involved and the services requested.  It could be anywhere from $200,000 to $2 million.  On the data servicing side, it will really depend on the size and complexity of the services requested.  There is a big difference in costs with a firm with 2,000 positions versus a firm with 2 million positions.

Michael Smith – we have advised on system implementation for managers with a few hundered million to half a trillion.  Most of the time these can be straightforward implementations but sometimes they cannot and that will affect cost.  The point of the systems are to minimize trade failures, make sure accounting is properly completed, 13Fs and 13Gs are not filed incorrectly and that understanding the various risks of the business.  These factors (along with reputation risk if something goes wrong) are weighed against the cost of implementation – for each manager it will be different in terms of ROI.

James Wu – after Lehman and AIG groups don’t realize how much exposure they have to a counterparty that is failing or about to fail.  Having that information is crucial for managers.

Ray Iler – yes, investors want to see that managers have thought about these issues and right now, in this tough fundraising environment, managers have to do what investors want.  If investors want full transparency, managers need to provide it.  Now, what about in-house technology solutions versus outsourcing.  Can managers do both?

Brian Lott – it is not a one size fits all solution.  It will depend on the risk profile of the manager.  What will happen many times is a sort of hybrid between the in-house team and our outsourced solution.  While the whole data warehouse will be a good solution, the in-house compliance manager or risk manager will usually want to have some kind of final sign off and control over the system.  Accordingly, many firms have their in-house team working directly with the data warehouse in a number of ways.

Joe Clark – the hybrid system is part of the philosophy that the system should match a manager’s specific needs.

[Some thoughts from James Wu and Michael Smith which I did not catch.]

Brian Lott – because each solution is tailored to each manager it is very important that stystem integration firms and the managers understand the scope of the work to be performed.  To do this the manager really needs to understand what the end goal is.  From the end goal the provider will be able to provide dates, an implementation plan, a description of the scope of work to be performed as well as a timeline of the phases of the project implementation.  It is very important for the manager and the provider to define the exact scope of the project.

At this point the floor was opened to questions.

Question from audience

  • What are some of the general needs of some of these managers (such as insurance companies, pension funds and family office managers)?
  • In light of the recent Galleon hedge fund insider trading case, do data warehouse solutions work to help  compliance personnel search emails to detect possible illegal activity.

Networking After Discussion

After the event we had the opportunity to do some networking.  I had to run fairly quickly, but I did have a chance to talk briefly with the following people:

  • Maital S. Rasmussen who works with small and start up hedge fund managers on their marketing materials and presentations.  Specifically I talked with Maital about presentation coaching which is becoming more important for managers who are presenting in front of institutional investors.
  • Erin Brodie who is the Senior Business Development Manager of Global Accounts at Advent Software.
  • Mason Snyder of Catalina Partners, which providers business risk advisory for institutional investments.
  • Maria Hall of M.D. Hall & Company Inc., a hedge fund audit firm.

****

Bart Mallon, Esq. of Cole-Frieman & Mallon LLP runs Hedge Fund Law Blog and can be reached directly at 415-868-5345.