One of the largest Margin Trading Product (“MTP”) brokers on the planet, DMM Securities of Japan, has announced their Australian subsidiary will stop taking and executing client orders around news events.  DMM is not specific exactly which news event will be excluded.  This is an incredible change of policy; news events create some of the highest-volume moments of each trading day / week / month.  This is like Target or Walmart deciding to close it’s stores on Black Friday here in the US.

The reason for this change can be boiled down to one of two reasons; DMM’s technology stack can not handle the massive spike in volumes around news events/releases, or their risk management practices are failing them.

I am not privy to DMM’s risk management practices.  They may be acting as an agency model and passing on all risk to aggregated counter parties which generates revenues by adding a mark up or a commission to each execution.  If this is their risk model, then the only reason to shut down trading during the busiest moments of the month is their technology stack can not process the tens of thousands of requests and execution confirms correctly.  Though they may be passing on all market risk to their own LP’s, the retail clients face DMM as the counter party.  And therefore DMM is likely on the hook for any technology failures that results in losses or missed potential gains from client orders/executions that are mishandled.  I am sure DMM’s account application documents state that DMM is not liable to their clients for any mishandled orders or failed technology; all brokerage firms include language as such.  Therefore if they are using an agency model, and still terminating trading during these news release moments, the technology failures can be expected to be massive.  Enough so that clients who have signed agreements absolving DMM of liability have strong cases that DMM’s technology failures are blatant, observable, and could have, SHOULD have been fixed.

The other possibility is that DMM operates on a principal basis and DMM has a dealing desk taking the other side of client orders and running a risk book on their desk.  (This post is not here to pass judgement on either an agency or principal revenue model; the most capital and human resource efficient brokers use a combination of agency and principal models).  If DMM is working on a principal basis and is turning off trading during news events, it is likely that their risk team has had large losses on news events.

I have worked as a dealer and as a risk manager on active desks.  And I can state from personal experience, and from the information I see on a weekly basis as a provider of technology to MTP brokerage firms around the globe, that news events are normally cash cows for principal-basis brokerage firms.  There are numerous reasons brokers tend to cash in during news events (spreads widen, over-leveraged clients react poorly, etc).  In fact I have seen risk desks have the most success during news events by applying a “hand’s off” rule (dealers can not execute hedging or prop trades for the 10 seconds prior to, and 20 seconds after a news release)  This “hand’s off” rule assumes the client flow will be fairly neutral and be executed over widened spreads.  The likely reasons for a principal-basis brokerage to lose money during news events are slow technology, or most likely, poor risk management practices.

The good news is that risk management practices can be changed quickly.  The bad news is a broker’s cost model of paying rebates to third parties for introducing order flow, fixed investment costs, or even direct variable costs are not easy to change.  So even though the risk/revenue mix can be changed quickly, the new revenue model may not support the cost structure of the brokerage.  Hopefully this change at DMM is a temporary change and the company can re-organize their risk management model in a way that still supports their cost model, yet allows for them to operate during the busiest time of the markets.

In July of 2007, the British Pound (“GBP”) vs the Japanese Yen (“JPY”) hit a price of 251.  That is, a single British Pound would get you 251 Japanese Yen.  It was a great time for UK residents to visit Japan.

By January of 2012, that same British Pound would only get you 119 Japanese Yen.  The relative value of the GBP vs the JPY had fallen be more than 50% in just four and a half years.  This time period saw numerous wealthy Japanese residents buying London properties with their (relatively) stronger Yen.

The macro forces that drove the GBP/JPY higher in the early and mid-2000’s was an expectation of rising short term interest rates in the UK (6% and trending higher), and the expectation for falling short term interest rates in Japan (0.3% and trending lower).  Essentially you could borrow cheaply in Japan and invest at higher rates in the UK.  This is what drove the currency pair to new highs.

And now we see the European Union’s Central Bank promising to fight deflation by lowering short term interest rates.  And on the other side of the world, the Bank of Japan is promising to do increase inflation rates in Japan by any means necessary.

So if the EU is actively dropping interest rates, and the BOJ is working to increase interest rates, is it possible that this year’s carry trade will be long JPY and short EUR?

Fred Wilson’s avc.com blog is must view material for me.  Fred is a part of Union Square Ventures and has a knack for seeing 2 – 4 years ahead.  His predictions for 2015 make for great conversation among those who pay attention to technology trends.

From Fred’s blog, http://www.avc.com / Jan 1, 2015

What Is Going To Happen