The Price Is Wrong
A cynic is a man “who knows the price of everything and the value of nothing” according to Oscar Wilde. But in 2007, as the global economy nose-dived and seemingly impregnable financial institutions crumbled into dust, more than a few commentators might have reflected that (and now I will paraphrase Meatloaf) “one out of two ain’t bad”.
Because as we witnessed billions being wiped off stock markets across the planet it became immediately clear that some of the world’s finest minds didn’t have a clue about either value or price, or anything in between. Financial assets had been systematically mis-priced until the house of cards could no longer be sustained - and apparently nobody had noticed.
Why? Simple: the dependencies and connections between various assets were unknown and unexplored, and so when individual links in the network failed the repercussions were unpredictable. The ‘price’ of a single security was derived without any real understanding of the constituent elements of that security and how they in turn were related to other financial assets. In short, the price was wrong.
Fast forward to 2017 in the UK and the collapse of Carillion - a services and outsourcing behemoth that specialised in public sector contracts. There the story was similar, so complex was the web of interdepencies and flow of revenue through the business that any attempt to accurately ‘price’ Carillion was doomed to failure.
The consequences were devastating, but there was one more twist in the tale: over two years later the various interested parties are still struggling to pull apart the tangled thread left behind after the collapse. The network was so complex, and the connections and dependencies so deep, that there was no easy way to identify and resolve all the challenges created when Carillion went under.
What Has Gone Wrong?
To answer that question we need to look very briefly at what we mean by ‘price’ and how central it is to the way our modern economy operates. A price is essentially a numerical value we place on something, either real or abstract, that balances both the benefits and risks associated with that entity.
In the context of financial markets, prices can usefully be thought of as a best guess between positive and negative outlooks, and incorporating a little of both to determine a fair middle ground. In the more specific example of securitised debt, the price reflects both the expected interest and repayments, but also the risk of default.
Of course in our largely capitalist economy, nobody creates prices - they are determined by the market acting on the best information out there. And they become the vital signal that all our actions are based upon. How and where we buy, spend and invest is ultimately determined by price - even if we are at times several steps removed from those decisions.
And as we noted above - the price of almost everything in today’s economy of ever-more complex interconnections is wrong.
That is a consequence of the way in which we view the world and the IT and data systems we have built to support that understanding. To take the first point - we tend to view the world as a collection of individual entities, or at least we do when thinking in terms of the business world. The entire science of economics is based upon this view of the world, one composed of ‘individual rational actors’.
As an extension of that viewpoint, we have tended to build IT systems, and databases in particular, to store and retrieve data on those individual entities. Think of a bank’s database of customers, each with their own name, address, balance, credit rating and so on. Fundamentally, we do this because it’s the obvious way to store and retrieve information that would be too hard to keep in our heads all day.
That last sentence provides a clue to the fatal flaw in this approach. It helps us retrieve facts, but it doesn’t help us understand the world as it is. And when we don’t understand the world as it is, terrible things happen. We need to re-think and re-purpose the way we store and interrogate data in order to deliver that understanding.
The Importance Of Connections
Connections define all of us, and prices that don’t truly understand connections and the complex web of interdependence around any entity are going to be incorrect. Take the example of a credit rating used above, but this time apply it to a business rather than an individual. This measure is a reasonable way of pricing the financial viability of an organisation, and all sorts of people, from investors, to suppliers and customers, are interested in that number (and want it to be accurate).
Connections define all of us, and prices that don’t truly understand connections and the complex web of interdependence around any entity are going to be incorrect
Conventionally we look at income, expenditure, balance sheets, and historical performance. That’s the ‘entity level approach’. But look deeper. The company in question (let’s call it company A) has a credit line with ‘company B’. A major customer (we will call them company C), responsible for over 30% of revenues, also has a credit line with ‘company B’. What happens when company B runs into trouble?
At the entity level - they lose a source of credit and have to go looking for another. At the relationship level, that challenge is compounded. A major customer has been similarly impacted. And this is - literally - the most simple possible example of this challenge. The real world is much more complex. Every single economic entity and asset on the planet is at the heart of a web of connections that is simply ignored when we attempt to price those assets. As a result, the price is wrong.
But we are trapped in a world built to analyse at the entity level. The vast superstructure we have built to do that job simply does not have the ability to map, follow and evaluate connections in the way that we need to. That needs to change.
Getting The Price Right
Resolving this dilemma isn’t just a nice to have. The global economy relies upon it. This is not an exaggeration. But it is worth acknowledging that any organisation or individual that can start accurately pricing is in an envious position today. Investors can make smarter decisions. Companies can evaluate the strength of their supply chain more effectively. Retailers can understand the true value of their customers.
So how do we make that happen?
The first and most important step is simple: build out a model of the real world and bring existing data into that model. To be a little more specific, this model could be described as ‘semantic’ - it attributes meaning to the data and most importantly defines the nature of relationships between entities and the rules that must apply to those relationships (which reflect the real-world rules that we are all familiar with).
When that happens, we have a world of connections and relationships that we can explore visually. More importantly, because we can follow these connections through multiple interdependent steps, we can calculate the real ‘price’ of any entity based on a true understanding of its location in a complex ecosystem.
When this model exists and connections are truly mapped out in a meaningful way, other possibilities then open up. We are able to ask open-ended questions such as:
- Are these two apparently independent entities connected to each other in some way?
- What is the true value of this customer, when taking into account both her previous spending pattern and network of influence?”
- What is the risk of default of this entity given its particular network of credit and supply relations?
These questions are at the heart of good business. But with conventional SQL based approaches, they simply cannot be answered other than in the most superficial manner and with vast amounts of processing power unavailable to all except a few. And even then, at a truly deep level, they cannot be answered at all.
With data management built around relationships and connections, they can. And as a consequence we can price effectively and enjoy the benefits.
December 13th, 2018
Share this post:
Answering The Really Big Questions
Posted March 25th, 2019 by Luke Feeney
Gartner and DataChemist Agree: Graph Databases Are Set For Growth
Posted March 4th, 2019 by Kevin Feeney