instagram arrow-down

About me

I founded and co-founded a couple of companies: Redington and mallowstreet; I write about issues of the day that touch me and make me think. Mostly about how to make things better.

Subscribe to my newsletter


Top Posts & Pages


Recent Comments


Arigata-meiwaku (verb)

Arigata-meiwaku is a Japanese word with no English equivalent. Which is not surprising, since it means: “an act someone does for you that you didn’t want to have them do and tried to avoid having them do, but they went ahead and did it anyway, determined to do you a favour. Then things went wrong and caused you a lot of trouble; yet in the end social convention obliged you to say Thank you”.

In the last few days, the commonly-used risk measurement, Value at Risk, (or VaR) has been given stick for performing arigata-meiwaku (or some variant of it) for JP Morgan’s loss-making credit derivatives special ops team (aka the Chief Investment Office).

Here goes:

The pundits are dissing VaR generally for not doing its job properly; namely, failing to sound the alarm at JP Morgan to warn of an impending mega-hit of circa GBP 4 Billion in its credit derivatives portfolio, when in reality it was never VaR’s job to work out whether JP Morgan could readily unwind its enormous and concentrated credit derivative positions and, if it couldn’t, then how much it would actually end up costing to unwind them in a highly nervous market which has just watched the painstakingly-negotiated Franco-Germanic austerity agreement, designed to stabilise Europe, getting unceremoniously ripped up by a new French presidential wildcard who then got vocal support from President Obama (Thanks, Big Guy) even as Greece prepares to vote on whether to abandon the Euro and, in effect, push Europe’s financial markets a few kilometres further towards the abyss!


Stay with me.

In those periods when history faithfully repeats itself, VaR is pretty good at predicting the severity of loss you might reasonably expect to suffer in your investments portfolio (at some given level of probability).

However, at times when history is pretty far from about to faithfully repeat itself (like right around now) and you have a huge illiquid set of credit-linked positions, VaR is really just doing you a favour and taking a stab in the dark on the size and arrival times of various industrial goods freight trains you should expect to come down the rail tracks across which you have tied yourself.


In fact, if you are a bank running a gigantic portfolio of concentrated credit derivatives in a wafer-thin market, VaR has very little interest in getting involved at all in your gig, and would much prefer that you use a couple of its sister risk measures, including the gorgeous, but brutally candid, “Scenario Testing”.

Scenario Testing is a What You See Is What You Get risk measure that doesn’t attempt to work out what is likely to happen on the basis of a complex statistical algorithm based on historical data. In other words, where VaR is stochastic, Scenario Testing is deterministic. It knows that in this fickle, crazy, messed up world economy we now find ourselves living in, it is close to pointless to attempt to calculate the difficulty (and, therefore, the expense) of unwinding your ocean-going credit derivatives portfolio in a hurry. So it doesn’t even go there.

Scenario Testing simply stresses market conditions to the max and says:


Let’s assume the market has discovered that you have sold protection the size of Alaska on an illiquid credit derivatives index (the CDX IG Series 9), and let’s also assume the market really doesn’t want to help you unwind your entire “bespoke” derivatives collection next Tuesday. In that super-stressed, low liquidity, Land of Mordor scenario, this is how much it would actually cost to close out your positions. Now, can you take that kind of pain? If not, don’t assemble that kind of portfolio!

Given the size and nature of its credit derivative positions, and given current adverse market conditions, that is the measure JP Morgan should have used to monitor and manage its risk. Maybe it did do some Scenario Testing, but maybe, also, it ignored the results until, eventually, the tempest grew too big for the teacup. Who knows?

VaR has its well-documented shortcomings, for sure, but on this occasion it must be rueing the day it ever agreed to dance the arigata-meiwaku. Things have gone wrong and caused a lot of trouble. And no-one is saying Thank you.

PS: Another genius word from the Japanese lexicon:

Age-otori (verb): “to look worse following a haircut”.

Now, fancy that! A Japanese word coined specially to describe Greece.

Love to get your thoughts

This site uses Akismet to reduce spam. Learn how your comment data is processed.