Luddites unite

I’m almost too tired to blog. Now you may ask “What on earth could leave Natan too tired to have an opinion?” Well, I have just finished reading a sixteen page interview with the principals at Themis Trading.

One again these self-proclaimed defenders of “fair markets” make dozens of claims about how exchanges, brokers and high-frequency traders are conniving to screw both retail and institutional investors. Here are some of my favourite excerpts:

• “But, let me be clear. We have no inside knowledge of these[HFT] firms. This is just what we hear in the market.”
  Dare I say that this could be a weakness in their position?
• “We have May 6 now to prove that HFT doesn’t increase market liquidity.”
  Strikes me that it also proved that the market isn’t particularly liquid when electronic market makers are forced out by stale data and unresponsive exchanges.
• “They provide it [liquidity] when they want to, not when the market needs them to. And only if their profit is virtually guaranteed… They are also liquidity demanders. Thesame guys who provide liquidity when they want to also demand liquidity when they need to. On May 6, they demanded liquidity.”
  Firstly, I imagine there are some HFT folks who will be delighted to know that their profits are virtually guaranteed. Secondly, what point are they making – that HFT firms trade for profit?
• “The basic problem, in our view, is the for profit exchange model, which is filled with inherent conflicts of interest… Traditionally, the exchange business wasn’t really very competitive, almost utility-like”
  Hang on, now I’m the one being blamed? They didn’t like exchanges when they were uncompetitive and slow, and they don’t like them competitive and fast. I know corrolation does not prove causation, but I think there might be an argument that a profit motive and competition amongst exchanges has spurred innovation, driven efficiencies and lower costs. I think their point is that because HFTs trade the most volume, exchanges are more likely to cater to their needs than to those of institutional brokers (or end investors) - which they support with...
• “Well, because we are not on the inside of these robots’ algorithms and their trading strategies to see exactly what’s going on, nor are we involved in the meetings in which we believe the exchanges are complicit in so much of what’s going on, it’s hard for us to come back with specifics when defenders of HFT say, “Oh, you don’t have the data to back it up.””
  So they’ve insulted HFTs, accused the exchanges of being complicit, what next – suggest that every other broker on the street is also involved in the great conspiracy?
• “Most institutional algos use a smart router to route orders in small pieces throughout the day. The pecking order of these routers differs depending on which broker sponsors the algo. But a common goal is to always route to the least expensive destination first. Most of the time this means routing to a dark pool before routing to a displayed liquidity venue."
  Aha, no surprise there then. Well, European MTF dark pools are more expensive than lit pools, so the argument that brokers use them to reduce costs doesn’t stack up. And in such a competitive environment, nor does the suggestion that the majority brokers act against their clients’ best interest – if that were true, Themis would be a large brokerage house rather than just “two or three guys”.
• On the OrderID and Side-of-aggressor data from dark pool data fees they say
  “By the way, they did get rid of them awfully quick overseas after we called attention to them. They were able, technologically, to do it in a heartbeat over there when some institutions started to boycott their European dark pools. Though, frankly, we’re a little skeptical that they took out everything we’d find objectionable if we had the regulatory power to comb through their records.”
  I’m sceptical they care about the truth – but it’s important to note that they don’t need any “regulatory power” – our public feeds are exactly that – public. So come and take a look.
• “Almost everyone else seems to have a vested interest”
  Really, I’m lost for words.
  

I guess an informed debate is too much to expect?

I'd welcome your comments...

Dangerous Opacity

Last week I blogged about the impact of electronic market making on lit books, concluding that whilst it has been positive, institutional investors sometimes need alternatives which allow trading with less market impact. We also believe in the power of competition to drive innovation, reduce costs, and make our industry more efficient.

Apparently, one of our largest competitors has reached a different conclusion.
Executives of NYSE Euronext have argued that competition has driven over-fragmentation(1), that established markets should not be able to use maker-taker pricing(2), and that alternatives to lit books are ‘ not legitimate’ and are causing ‘dangerous opacity’ that will undermine price formation and confidence in our markets(1).

What is motivating these arguments?

Does NYSE Euronext really believe that fragmentation can go or has gone too far?

• One feature of US markets that really interests me is that most market operators operate multiple order books with differing tariff structures – fragmenting their own market so as to address different customer segements. NYSE was amongst the first to do so; it operates its hybrid electronic/human Classic (floor) market and also the fully electronic NYSE Arca – and does not seem about to merge them together. So they apparently have no problem with contributing to fragmentation in the US.
• My opinion is that European investors, brokers, market operators and regulators are not yet fully accustomed to competition and fragmentation. But, having worked in New York when ECNs first proliferated, I’m not worried. It will take some time, but wider adoption of Smart Routing, more efficient clearing arrangements, further standardisation of tick sizes and volatility interruptions, and consolidation amongst venues will make today’s concerns about fragmentation seem quaint.

Does NYSE Euronext really believe that use of maker-taker tariff models by established venues distorts the market?

• Apparently not - NYSE Arca in the US operates maker-taker pricing, and has done for years. As for Europe, it’s difficult to comment, since details of the exact pricing incentives available to ‘liquidity providers’ are not readily available on their website.

And does NYSE Euronext really believe that everything except block trading should be completely transparent so as to avoid ‘dangerous opacity’?

• It doesn’t seem so to a casual observer. The NYSE Classic market (the ‘floor’) in particular stands out for having an unusually opaque model. By according them what they refer to as “parity”, certain members get to jump the queue completely;
  


• Orders from Designated Market Makers (rebranded ‘specialists’) get parity with other participants’ orders. This means that if there are 5,000 shares on the bid from a number of ‘normal’ members, and the DMM subsequently bids for 500 shares, the first 1,000 shares of an incoming sell order will be split 50/50 between the DMM and the other queued limit orders. DMM’s get this privilege as compensation for their obligations to maintain a “fair and orderly” market – though given the amount of money DMM priviledges change hands for, parity clearly has significant economic value.
  
• Similarly, each individual ‘Trading Floor Broker’ also gets parity with normal members and with DMMs, and can jump the queue without even needing to display their orders publically. They also get unique visibility of, and access to, market-depth data that is not visible to normal members. I’m not sure they have any obligations, but giving parity to floor brokers allows them to attract order flow and hence ensures that the floor looks like a hive of activity on television.
  


• Personally, I don’t like it when people cut in front of me in a queue, and given a choice, I wouldn’t choose to stand in a queue if it was specifically designed to work that way – although I welcome the fact that particpants are offered the choice. But, in my mind, NYSE Euronext’s commitment to this model does rather undermine their argument that all non-block trading should be fully pre-trade transparent to avoid ‘dangerous opacity’.
  

A lot is at stake in the MiFID review – and the debate is especially heated around the topic of non-displayed trading – both in MTF Midpoint Books, and Broker Crossing/Internalisation Systems. The exchanges that pre-MiFID enjoyed the protection of “concentration rules” are struggling to come to terms with a landscape in which they face competition – both from market operators and (to some extent) from their clients. So they try to persuade politicians, regulators and the general public that we’re approaching a precipice and need to turn back, arguing that whilst MiFID was intended to unleash competition and choice for market participants, the exchanges were supposed to win.

At Turquoise we believe that;

• Transparent price-time markets are not ideal for every order. There is a legitimate market need for alternatives to lit orders books.
• Competition is somewhat distorted and price formation potentially affected by inconsistencies in the rules applicable to MTF and broker-operated non-display facilities. But, rather than trying reduce choice for investors, a better solution would be improved post-trade transparency (to avert worries about price formation) and a relaxation of the waivers that limit innovation (and thus competitiveness relative to broker crossing systems) of Regulated Markets and MTF-operated non-display venues.
• Concentration rules were eliminated because they had proven to be a barrier to competition and innovation. Brokers now have greater flexibility to internalise, and whilst in theory that could ultimately lead to lower volumes or wider spreads in public limit order books, there is no evidence to suggest either US or European markets are close to such a scenario in reality.

Slow down, you move too fast...

Before tackling some of the potential downsides, here’s why speed and capacity are important:

• Market confidence suffers most when participants worry that they cannot trust the prices they see, or that they have lost control of their orders. Capacity and speed are essential to maintaining the confidence of investors, without which, liquidity suddenly evaporates (as it did in US markets on May 6th).
• The guarantee that markets will be highly responsive, and that an order can be amended or cancelled at any time, gives participants the confidence to expose limit orders that they might otherwise withhold from the market.
• The ability to execute instantaneously across multiple markets, instruments and asset classes leads to less risk for arbitrageurs, resulting in greater market efficiency of correlated assets. This reduces hedging costs and encourages liquidity provision and capital commitment.
• Markets with inadequate capacity or throughput reject orders at busy times (either implicitly or explicitly), reducing the liquidity that might otherwise be available just when it’s needed the most.

For the above reasons, greater speed and capacity drive narrower spreads, which reduce overall transaction costs, improve overall investment returns, and help companies raise capital more cheaply (helping the real economy grow).

But what about relative speed and relative costs – are long term investors being systematically disadvantaged by those trading faster and at higher frequency? Do faster markets benefit one group of participants more than, or even at the expense of, others?

I think there’s irrefutable evidence that competition amongst markets and amongst electronic market makers has resulted in a dramatic narrowing of spreads. Unambiguously, tighter spreads mean improved execution quality for retail orders (the majority of which are marketable). So it’s hard to think of the ‘amateur’ market participants as victims in our new competitive markets.

What about institutional investors with large orders? Are market professionals the victims?

• Would markets be better without electronic market making (EMM) firms?
  No. We know exactly what markets look like when EMM firms either cannot or choose not to participate. EMM firms operate with extremely low margins, and are incredibly sensitive to risk – so they’re the first to be forced out of the market when systems slow down. May 6th was instructive – when EMMs back away, volatility increases dangerously.
• Is the liquidity provided by EMM firms ‘real’ or ‘ephemeral’? If you take liquidity from an EMM that immediately unwinds for a profit, would you have been better off not taking their liquidity in the first place? If they end the day with a flat book, but have made money, has that come out of the institution’s pocket?
  In price-time markets, you always get the best available price at the time. Liquidity offered by EMM firms is just as real as that offered by any other participant – except it’s often more competitively priced. And if the EMM firm with which you trade is able to profitably unwind that risk in other stocks or asset classes, rather than by directly covering the position, then you’ve accessed liquidity that wouldn’t have otherwise been available. Their profit is not necessarily your loss.
• Is it fair that some participants invest in the fastest technology and most sophisticated algorithms, and co-locate their systems beside exchanges, potentially allowing them to react more quickly than other participants?
  Fair or not, it’s absurd to imagine that we can effectively legislate against profit-maximising behaviour by market participants. If co-location was prohibited, then firms will instead congregate in data-centres adjacent to the markets. And there’s no way to ensure that all participants receive all data simultaneously – even if we had a consolidated EBBO (which I’ve argued against previously). To borrow a phrase from politics – it’s about “equality of opportunity” – we have to recognize how much fairer and more efficient markets are now we have a level playing field with many firms competing to add or remove liquidity than they were with ‘designated’ specialists or market makers (who enjoyed special privileges).
• Do some automated trading firms exacerbate volatility or engage in ‘momentum ignition’ by stepping ahead of large orders, forcing them to revise their buys upwards and sells downwards?
  Here’s the rub. Efficient markets are supposed to ensure that the price “reflects all available information” – and automated trading firms specialise in recycling the information represented by market data back into the price. If the market is aware of a large buying or selling interest, it’s only natural that prices move – with “market impact” and “market efficiency” being closely related. For such trading strategies to be viable, automated traders rely on the propensity of brokers executing client orders to “chase” a stock up or down regardless of “fair value”. Whilst this all within the rules (let’s trust that market surveillance is effective in spotting rule-breaking), it can be frustrating and costly for a trader who sees their order being stepped ahead of.

So there might be some scenarios in which automated traders take advantage of institutional flow – but what, if anything should we do about it? Is it a matter of brokers becoming more sophisticated, or clients changing their trading style? Or do we need regulatory intervention to protect one category of professional participant from another?

Two thoughts spring to mind:

• Just as most financial professionals oppose protectionism in the real economy, we shouldn’t ask or expect regulators to make financial markets ‘safe’. It’s the competition amongst market participants that drives market efficiency forwards and encourages brokers to invest in smarter trading strategies.
• Any regulatory ‘cure’ (such as a transaction tax, or other constraints on trading activities) would be far worse than the problem they’re trying to fix – ultimately sapping liquidity from markets and driving spreads and volatility higher, to the detriment of institutional and retail participants alike.

The real conclusion to draw is that transparent price-time markets aren’t ideal for every order, and so brokers and market operators need the flexibility to offer investors alternative solutions that allow them to access liquidity in the manner most appropriate to the order in question. Given individual investor’s preference to trade with less market impact, post-trade transparency is essential to ensure these alternatives still support efficient price formation.