How Big is the Haystack?

At DealVector, we often say that our mission is to help one needle in the haystack connect with another needle in the haystack, without alerting the whole haystack. The haystack is made up of all of the holders of financial instruments. So, how big is the haystack?

There are several kinds of assumptions we can make in answering this question. McKinsey puts the total global stock of financial instruments (securities, loans, etc.) at something over $200 trillion face value. All global derivatives together have a much larger notional value – one that is variously estimated as sitting somewhere between $600 and $700 trillion.  For context, that last figure is roughly 10x global GDP.

But the need to communicate varies across asset classes. Greater urgency exists in cases where liquidity or visibility is poor, or where control issues arise. So in our rough cut at the size of the DealVector haystack we toss out several big categories: equities, most derivatives, and most government securities. What we are left with is primarily credit-related products, as well as fund or LP interests. Credit includes straight corporate bonds, especially when it pertains to distressed or work-out candidates. But we focus particularly on CDOs, CLOs, RMBS, TRUPS, and any tranched asset class. The total size of the haystack estimated in this more narrow way still adds up to about $100 trillion, or more than 1x global GDP.

So that may be the size of the haystack. But how many needles are in it? How many holders are there of particular financial interests?  Well, from some of the LP data we have seen, there is a very long tail to the distribution of holders. Large institutional holders possess roughly 80% of LP fund interests, calculated by the value of those interests. But High Net Worth holders, relatively smaller, control about 95% of those LP fund interests, calculated by the number of those interests. What if we use this distribution of LP interests as a model for all of our addressable asset classes? So, institutional investors plausibly control $80 trillion out of our $100 trillion target set, but they constitute only 5% of the total number of positions held.

How many positions might this be? In North America, a reasonable range of the number of distinct financial instruments applicable to DealVector is between 5 million and 10 million total instruments. North America, in turn, constitutes about 1/3 of the global applicable deals. So, call it between 20 to 25 million instruments.

Are there 10 holders per deal? 20? 30? Much more? The number will vary significantly by deal type. On many deals there will be hundreds of holders. In others, like the AAA portion of a CDO, there might be only 1 or 2.  If we keep the math simple (as well as probably conservative), we might assume there are 20. This would still imply something like 500 million distinct position holdings globally for our targeted asset classes.

Let’s close with a thought experiment. If big institutions hold 5% of the 500 million, then that would imply 25 million distinct institutional positions. Each holder would have an interest in connecting with the 19 other (assumed) holders of its deal. It would also have an interest in connecting with holders of similar deals, which would be a multiple of the first deal.  So now we are talking about billions of potential connections – connections that might be beneficial to both sides, but which have been impossible to make under legacy systems. That’s a lot of hay.