KNIGHT VISION.

Brendan McCarthy, Knight Direct’s head of sales and relationship management talks to Best Execution about recovery from the nearly disastrous glitch of August 1st, and the lessons learned as well as the impact of impending regulations in the US and Europe.

What has been the impact of the technology problems that cost the firm to suffer a $440m loss in August?

The truth is nothing is perfect and there are errors. If you look at the stock market crash in the 1930s, it was a fully manual market that was responsible. In today’s electronic markets, humans still make mistakes. Following that mistake, we received

very strong support from both clients and our competitors. Three investors (Fred Tomczyk, CEO of TD Ameritrade, Martin Brand, managing director of Blackstone and Matthew Nimetz, advisory director of General Atlantic), have joined the board in the past month. At Knight Direct, our technology and business of providing agency algorithms is separate from the affected unit, so we were able to bounce back quickly once it was clear there was no impact on us. Across the entire firm, Knight initially lost market share, but the flow has returned and we have regained our leading position.

What accounts for the support?

I think it is a combination of culture, products, research and client services. It is also down to the way we handled ourselves. We have good relationships and have always put the client first. However, we also were transparent about the error and have learnt our lessons. We recently hired IBM to look into the trading glitch and are currently conducting an internal and external review. We are also putting in place better policies, practices and procedures to avoid this from happening again.

Looking at the market as whole, what impact do you think the impending regulation will have?

I think the competition that has been created by regulation such as MiFID in Europe and RegNMS has led to innovation. We have seen a plethora of alternative trading platforms and new solutions such as smart order routing and sophisticated algos to source that liquidity. In turn, the primary exchanges have lost market share to off- exchange venues. This is not the case in Europe although we are seeing an increase in terms of the concentration of flow to these off-exchange platforms.

Do you think the rules have made it easier to achieve best execution?

The execution experience has materially improved and the average execution speed has dropped significantly to less than one second compared to north of three to four seconds in the past. Innovation has also meant that the costs of trading have been dramatically reduced for the end investor.

Where are we today?

The European framework is more nascent than in the US and is guided by a different structural framework. It is not as efficient or cost effective as the US because of the clearing and settlement costs. Central clearinghouses (CCPs) need to interoperate in order for the European market to develop further. What is happening is that many are trading American Depositary Receipts (ADRs) instead of directly on the platforms because it is less expensive.

Do you see interoperability happening?

It depends on where MiFID goes. If the regulation can achieve interoperability then the implicit as well as explicit costs will decrease and it will create a much better bid/offer spread. It is not clear though that this will be the case with the current form of MiFID II. At the moment, everything is being proposed, but nothing is being enacted. If it does not happen then it will be difficult for the European market to get to the next level. This does not mean that interoperability will go into reverse but that it will not reach US levels.

In terms of tools, what are clients looking for?

In the US, there is clear need for intelligent sourcing algos. The development of effective algorithms requires heavy investment in technology, market data and strategies to extract alpha as well as connectivity. We believe the value of a broker like Knight is to help clients efficiently navigate the market through sophisticated trading tools that lower the market impact as well as explicit and implicit costs. We can offer customisation in that we can take a standard algo and configure it to better meet a client’s particular trading strategy. But more than that, we design and improve our algorithms on in-depth research, that’s our differentiator.

There are concerns about high frequency trading. Do you think it is warranted?

There are four different types of high frequency traders – passive, stat arb, directional and structural. The majority are passive and provide some kind of liquidity but there are those that act very aggressively and there needs to be some kind of regulation around that particular activity. They should not though all be lumped together.

What are your plans for the future?

We plan to continue to invest heavily in our algo research and development of new, more intelligent tools for our buyside clients. We want to optimise their trading strategies and level the playing field for all market participants.