• 04 November, 2008 12:46 PM EST
• What Would Markowitz Say?
• Posted By: Robert Handler, Research VP

I read a recent Wall Street Journal article entitled "The Father of Portfolio Theory on the Crisis," where they interviewed Harry Markowitz, Nobel prize winning father of modern portfolio theory, about the current financial crisis. Markowitz greatly influenced what I put in "IT Portfolio Management: Unlocking The Business Value Of Technology." The article also directed me to an essay by Harry Markowitz posted on the Index Funds Advisors Web site titled "What to Do About the Financial Transparency Crisis." The three main messages I got were:

1. Investments should have uncorrelated risk
2. Transparency of investments is critical to understanding them and will be critical to understanding the magnitude of the recovery and subsequently recover.
3. "Selling people what buyers and sellers don't understand is not a good thing"

Having co-authored a book on IT portfolio management, I wondered if it were possible to apply these three main messages to IT portfolio management. I believe it is and will try to do so in this blog entry.

First, I'm fairly convinced that, on aggregate, we do a horrible job of understanding the risks of our IT investments. Without understanding the risks, we fail to understand the risk correlation. While I'm not suggesting IT organizations run their project portfolios through SPSS to verify correlation, it would be wise to spot check individual and aggregate portfolio risk. Do reasonable risk ratings exist for your IT investments? Are they spread across different investment categories? Different business units? Different "bets" as one client likes to call them? This would be wise to check, now more than ever.

Second, Markowitz advocated transparency, both as a preventative measure and an aid in recovery. Organizations can and should be more transparent with their IT portfolios. What does all of this stuff cost? It's expensive. What's the risk that it won't perform or complete as expected? It might be high or low. Either could be ok, depending upon your investment objectives and risk tolerance. If there is no transparency, however, it is difficult to manage expectations, course correct, and provide quality service.

Finally, Markowitz stated "selling people what buyers and sellers don't understand is not a good thing." I think this is an especially salient point in IT portfolio management. Have you ever heard someone float an idea for an IT project that made absolutely no sense? What is even worse is when groupthink ensues and folks pretend an IT project makes sense even when no one really has a clue what it's about. This should not happen. It does far too often. You should not need rocket scientists to decode business cases - they should be self explanatory and rational. This may, in fact, be the most important point. If a project doesn't make sense, don't do it. If it made sense at one time and no longer does, stop doing it.

So, in a nutshell, when it comes to IT portfolio management being applied to projects, applications, or whatever, it looks like a few basic concepts can be borrowed from the investment world. First, manage risk. Not just individual risk, but aggregate risk - manage dependencies. Second, demand transparency. It will foster understanding and maintain proper expectations, possibly even preventing calamity. Finally, and at all costs, avoid doing things that don't make sense. If they don't make sense to you, they probably don't make sense to others as well. Stand your ground.

Submit a Comment | Permalink

Comments

• 17 December, 2008 12:46 PM EST
• Tom Wright wrote:

I too have given thought to MPT principles and their application to IT Assets, but my conclusions were slightly different.

Clearly MPT is about balancing risks and returns on an efficient frontier by combining risky assets with non-risky assets to improve overall portfolio performance. Furthermore, I believe MPT combined with Efficient Market Hypothesis/Theory (semi-strong) adds another critical dimension in the application of MPT principles.

My thinking is that the market portfolio (free from company-specific risk), is represented by market-available software. Whereas the most common, or most commoditized business processes (finance, HR, Treasury...) are captured in the market and possess the least risk (and one that is constantly reduced over time). Following that, best of breed solutions (again, market-based) offer more industry-specific efficiencies (container shipping, warehousing, ...) and lastly, those business processes that offer competitive advantage (akin to "insider knowledger" in semi-strong version of EMT) is developed internally.

On the efficient frontier, you start out with low risk, broadly defined, common business process, with low overall return, represented by ERP software, as we move to the right on the efficient frontier, we have best of breed software that is slightly more risky, but with better returns and slight differentiation. On the far right, highest returns, highest risk is the in-house development, and perhaps niche development (which should be replaced by commodity software as soon as it is available).

I think using this as a template reduces overall IT investment risk as well as assures minimally accepted productivity in non-differentiating business processes that is market-based. It also outlines areas of investment on which it is best to focus high-end IT investments where the potential returns warrant the risk.