Platt Perspective on Business and Technology

Deterministic, semi-deterministic and stochastic business models, or fun with business numbers

Posted in macroeconomics by Timothy Platt on October 31, 2013

I first started thinking about the issues that I will discuss here over a year ago. I had them in the back of my mind and kept coming back to them to set them aside again until later. And two days ago, as of this writing, I finally decided that “later” had finally arrived. And yes, I am writing this to go live on Halloween day, which is perhaps appropriate given the tenor and direction of my intended discussion here. So think of this as a business world Halloween offering, that is hopefully more treat than trick. And I begin this with the fundamentals.

I have written repeatedly, throughout the span of this blog about business models, and most of what I write about here revolves around them too. But I begin my discussion here in very abstract terms. I have written about:

• Virtual and outsourced business models (see for example my series on that approach at Business Strategy and Operations, postings 127-129),
• Open business models (e.g. my series on that approach in that same directory as postings 130 and loosely following for its Parts 1-5),
• Globally-expansive business models (e.g. Globalization and the Business Model Paradigm,
• Online business models (e.g. my series: Opening Up the Online Business Model for New and Emerging Opportunity, at Business Strategy and Operations – 2),
• Dynamic performance based business models (e.g. my series on that paradigm at Startups and Early Stage Businesses, postings 123 and loosely following for Parts 1-8),
• And more.

I step back from the contextual specificity of these postings and series here. And I begin that my noting that:

• In the most abstract sense an effective business model is the coherent, organized presentation of a system of relationships between a set of factors that collectively represent that business and its operations and strategy, and that encompasses its overall goals.
• The system of relationships between these factors, generally statable as a set of logical rules, serves to set the basic parameters that this organization will be build and run within, and for both its ongoing goals and priorities and for determining what specific operations and activities would and would not fit into this business as making sense for it. In this, as a point of clarification, following a business model and building and running an organization in terms of one is often in practice, at least as much about knowing what not to do as it is about knowing what to do and about knowing when something that should be done, should only be done as meeting a low priority requirement.

Now let’s consider those factors that enter into these business models:

• Some of necessity are going to be categorical and non-numerical in nature,
• Some are semi-quantitative,
• Some and perhaps most are quantifiable, and of those some are set value constants, and
• Some are variables. And of those factors, some present themselves as deterministic, where a given calculation involving them can only legitimately assign one set value to them at any given time and under any given set of calculated relationship circumstances.
• Some of these factors present themselves as stochastic, and their value varies within a perhaps bounded range, but specific values for them will vary within that range according to a knowable statistical distribution.
• Of these values, some change in a seemingly random manner but within predictable bounds,
• Some shift along trend lines (where range boundaries can shift with time),
• And some at least offer potential for breaking outside of any predictable range and that is where businesses have to accommodate the unexpected and perhaps unpredictable.

So far this is probably abstract enough to qualify as trick. But my goal for the balance of this posting is to bring this discussion far enough from the entirely abstract so that hopefully it will end up on the treat side of that balance. And I begin with our major financial institutions going into what rapidly became the Great Recession. And more specifically, I begin with a focus on the very cores of their business models where they would strategically and operationally evaluate the potential risks and benefits of the investment instruments and other monetizable resources that they offered, and that they accepted as collateral.

• If you assume a key set of operational relationships that underlie your business model are fixed and deterministic, or that they might vary statistically but only within narrow and controllably fixed, predictable ranges – but they veer outside of your allowed-for ranges, this creates problems.
• If you build operational systems that create risk/benefits balances that you do not fully understand, as was the case with many of the complex derivatives and other investments that so many of these institutions generated, and that was the case across the board for subprime mortgage lending, then you court inevitable disaster. And for these institutions, inevitable happened.

The leaders and managers of these institutions acted as if all of their business models were fully deterministic, and when they stepped back from that level of assumed certainty, it was to entertain the possibility of variance within readily contained and limited ranges. Ultimately they were only looking out for their own personal wealth and their own personal financial security, but even there, they failed to see that if their businesses that created this wealth for them were to suffer, they ultimately would too.

I named this posting “fun with business numbers” because at least in retrospect, too much of our overall economy was in effect being played like a game – and was being gamed outside of their own declared rules for that too, in the process – and with inevitable results given the fundamental misunderstandings in place as to the nature of those factors they were dealing with or the ranges that their values might shift into and beyond.

And here and in this context I offer “semi-quantitative”, or if you will “semi-deterministic” as if ultimate Halloween business tricks. Making numerical calculations where they would seem easier and more favorable in predicted outcome, and simply leaving the rest untested and treating that part of their systems as if entirely categorical and descriptive was and is a formula for disaster. And I write this fully aware of the mathematical sophistication applied by quantitative analysts (or quants) at these institutions to the investments and investment portfolios that they modeled and assembled. You can apply the most sophisticated and insightful mathematical reasoning to a context and still be fundamentally wrong and misleading, and even when you are seeking clarity and certainty – if you start with faulty assumptions. And the faulty assumptions that these institutions and their experts made cut to the core of all of their reasoning, starting with misunderstanding the basic factors that they wished to model and correlate as they built and followed their basic business models.

OK, the second half of this posting is mostly trick too. What were the odds of that happening? Have a treat-filled coming year. And meanwhile, you can find this and related postings at Macroeconomics and Business.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: