EVOLUTION OR REVOLUTION?
By Paul Reynolds, CEO of BondCube.
The evolution of fixed income trading technology is not often described as steady. There was a burst of activity fifteen years ago and now there is a further frenzy of activity to try to unlock the present liquidity conundrum.
Fixed income trading technology is trapped in a triple clamp of huge regulatory change, sub-optimal market structure and pinched sell-side budgets. It is not all bad news though, as electronic trading volumes have in fact risen, just not as much as you might expect compared to the overall growth in size of the market of 50%, and the growth is mainly in small orders.
Hidden in these statistics lie the underlying problems of the evolution of fixed income trading technology, or you might say, the lack of consistent evolution of fixed income trading technology, compared to other asset classes like equities.
Fifteen years ago companies like Bloomberg, MarketAxess and Tradeweb introduced ground-breaking technology that provided the channel for investors to ask multiple banks for prices to trade, known as request for quote (RFQ). These companies now dominate corporate bond trading in the US and the EU.
After fifteen years RFQ still remains the only effective electronic method of trading between investors and banks. Before the crisis it was possible to use RFQ to execute all sizes of orders. Since then order sizes have consistently declined, although the number of them has grown dramatically.
So what about the large orders? Investors have an ever-increasing backlog of unexecuted, large orders known as ‘failed to trade’ (FTT). If fixed income is to evolve it must address how to improve the efficiency and cost of executing large orders.
The first talon of the triple clamp is the new regulations requiring banks to apply more capital to conduct fixed income activity with investor clients. The result is that their capacity to provide buying and selling prices to investors has fallen by some 75% since the 2008 global financial crisis.
RFQ relies very heavily on banks providing accurate pre-trade prices followed by execution at those prices. Any investor will tell you this is simply not the case these days.
The second talon is sub-optimal market structure. Existing platforms have tried new variations on the RFQ model or asked investors to submit entire orders to banks to find matches. The evidence suggests none of them has worked so far.
Finally, banks have recently been through a cycle of single dealer platform initiatives that did not deliver because they further fragmented liquidity. Scarce IT budgets have led them to consider third party solutions to return their trading and sales businesses to profitability.
So, at a time when investors have more exposure to bonds than ever before, at historically low interest rates, they are relying on constrained bank liquidity within an RFQ model that struggles to execute large and illiquid trades. This concentration of liquidity risk amongst investors has recently caught the attention of the Fed and the SEC.
This situation is unique to fixed Income. In other markets like equities, FX & futures everyone trades with each other through exchanges and volumes are huge, especially when there is volatility. In fixed income volatility means dramatic price changes and very low volumes as recently witnessed in the high yield market.
This is because banks are solely responsible for setting prices and deciding which volumes trade. If they do not want to buy in a falling market then very little investor selling will take place and prices will be marked down. Even if a buyer wanted to take advantage of lower prices they would struggle to find the aforementioned offers as there is no effective way to discover them.
With banks unable to ‘grease the wheels’ the market structure needs to evolve in a number of ways to connect buyers with sellers outside the RFQ channel. The first way is to emulate the other markets and permit ‘all to all’ connectivity.
So what would this ‘all to all’ infrastructure look like and would it in fact be like equities, FX & futures? The main difference between these markets and fixed income is that in fixed income there are hundreds of thousands of different bonds that trade very infrequently. In the other markets there are just a few that trade very intensively.
All to all would certainly be a significant evolutionary step, but can technology innovate sufficiently to overcome the different make-up of fixed income compared to the other markets, to make it work?
If you have a large trade to execute in fixed income you need to be very careful who you tell. The only person you really want to tell is a person who wants to do the other side of your trade, so if you are a seller, you only want to find a buyer and vice versa, not another seller or even someone who has nothing to do. They may use the knowledge that you are a seller to their own advantage and your disadvantage.
So connecting everyone together using ‘all to all’ is one thing, but minimising ‘information leakage’ is another. Posting enquiries by deploying ‘indications of interest (IoIs) with just ‘side’ and ‘minimum size’, but no price, significantly reduces market risk. Adding ‘dark matching’ is however the key to ensuring minimal market risk.
Dark matching is where a seller can select a group of other users to ask who is a buyer, but the only person who will actually know, is a genuine buyer, because they have entered an enquiry asking for a seller. The two matched users can then negotiate and trade privately, posting the traded volume afterwards.
Matching a genuine buyer with a genuine seller also means the transaction costs should be much better. This is because the buyer and seller are likely to negotiate a price between the bid and ask price quoted in the market. Neither has to hit the bid, nor lift the offer.
Fortunately for fixed income there is abundant effort and activity in introducing new ideas to solve the failed to trade problem, it remains to be seen if the answer will be a giant leap or a steady evolution.
© BestExecution 2014