August 06, 2019 Speed Bumps & the Transaction Fee Pilot: Two Wrongs Don’t Make a Right Larry Tabb TABB Group Follow | Profile | More Share Though moving forward with an asymmetric speed bump as a liquidity provider incentive may somewhat cancel out the incentives lost by banning rebates, two wrongs don’t make a right. Both of these initiatives would be unhealthy for the markets, argues TABB Group founder and research chairman Larry Tabb. Last month I penned an SEC comment letter on Cboe’s proposed EDGA asymmetric speed bump (see: “Speed Bumps vs. Maker-Taker: The Lesser of Two Evils?” July 17). Though I clearly stated that I was not a fan of speed bumps in any color or flavor – whether they are non-denominational bumps that slow both liquidity providers and takers or discriminative bumps that favor one or the other – I also stated that if the SEC bans rebates, this speed bump may provide a mechanism to incentivize market makers to continue to populate trading books. While I still agree with both of these statements, I would like to categorically state that I believe both of these initiatives are unhealthy for the markets. Though moving forward with an asymmetric speed bump as a liquidity provider incentive may somewhat cancel out the incentives lost by banning rebates, as we were told many times as kids: two wrongs don’t make a right. To rehash a bit … The SEC is in the process of implementing a transaction fee pilot that is intended to test the impact on order routing of the 30 cents-per-one-hundred-shares (30 mils) Regulation NMS access fee cap and exchange rebates. The pilot will create two additional test tiers: one pilot tier will drop the 30-mil fee cap to 10 mils, while the other will eliminate the fee cap but ban rebates. In my comment letter, I stated that, in general, it is beneficial to have liquidity providers/market makers compete to populate the order book with generally accessible and transparent liquidity. That is the crux of our markets. Without firm, transparent liquidity, we wouldn’t have the high-quality capital markets that we have today. Historically, exchanges have incentivized liquidity providers/market makers with either an information advantage or a cost advantage. We eliminated the information advantage more than a decade ago and moved toward a cost advantage under Reg NMS, which is when the 30 mil access fee cap was implemented. The newly proposed Cboe asymmetric speed bump is a different way to provide market makers and liquidity providers with an incentive, I argued. This new Cboe-defined “Liquidity Provider Protection” capability is intended to allow market makers to quote more aggressively, given that they will have 4 milliseconds to decide whether to cancel or move their possibly overly aggressive quotes before they are executed. Besides arguing that all speed bumps are detrimental from a market structure perspective, I argued that, if the SEC decides to move forward in banning rebates, maybe this new asymmetric speed bump would provide an incentive for market makers to populate exchange-type, transparent trading books, even though slowing liquidity takers via an asymmetric speed bump does the following things: Creates an advantage for sophisticated liquidity providers over non-sophisticated market participants; Introduces a systematic fading option for sophisticated traders (somewhat akin to the widely derided “last-look” practice in FX /OTC markets) to our exchange-traded markets that have been built on a foundation of reliable, firm, and immediately accessible quotes; Legitimizes fading (or, you could say, “spoofing”), as purportedly accessible quotes would not be hittable; Switches incentives from rewarding trading to rewarding not trading – a rebate is earned only when you do trade, while the asymmetric speed bump enables aggressive traders to duck out of the way of certain incoming orders and not trade; and Is a transfer subsidy from liquidity takers (i.e., most investors) to liquidity providers (which mostly are sophisticated technology-enabled traders). While my comment letter pointed out the challenges created by this new speed bump, I surmised that maybe this type of speed bump wouldn’t be so bad if the SEC were to completely ban rebates. While I think my opinion was pretty transparent, I need to make a more definitive statement: that not only am I in favor of the current incentive programs offered by exchanges (i.e., rebates), I also am against the asymmetric speed bump. And in my view, banning rebates and green-lighting an asymmetric speed bump, to me, are two wrongs, and two wrongs do not make a right. While many large investors and some brokers support the elimination of rebates – as they believe rebates can create conflicts of interest for agency brokers that, if not held accountable, will route to venues based upon what is best for the broker and not the client – I am not in favor of tinkering with current market structure. The current speed-aligned, rebate model has allowed retail and institutional investors to increasingly reduce their trading costs, putting millions (if not billions) of dollars back into the pockets of investors. For those worried about routing conflicts, work with your brokers to receive and understand the new Institutional 606 Reports; and if those reports do not provide the level of granularity you need to hold your brokers accountable for their execution quality, invest in venue and routing analytics and analyze the data yourself or with an independent provider. While competing on speed seems to be an increasingly expensive and losing battle, the current rebate model provides incentives to trade at aggressive prices, as opposed to ducking out of the way when the market moves. In addition, any trades in ATSs or over the counter, which are priced off the best bid-offer and/or the midpoint (approximately 38% of the market), receive better pricing – without having to jostle with exchanges’ top-of-book speed-dominated market. As I have stated many times, the current market, even with (despite) all of the insane speed and pricing structures (maker/taker, taker/maker, fee/fee, dark pools, crossing networks and algorithms), enables traders to determine the market impact and information leakage they want to pay for by using the different pricing models to fine-tune where in the market-wide liquidity book they want to be. If they are looking at cost, they can post at the high rebate exchanges; if they want less adverse selection, they can take a lower rebate to trade higher up in the queue. Now to do this, you do need to be astute, accurately measure how your brokers are executing on your behalf, and hold your them accountable for their execution quality. Do I believe that the current Reg NMS-defined 30-mil fee cap is perfect? No. What technology-based service has held price for 15 years? None. So 30 mils is probably too high a fee. But I don’t know what the right number should be. That said, most of the rebates supported by these fees find their way back to the client, either through lower commissions, price improvement, or better service, as our US equity markets are incredibly efficient. And by the way, what broker, whether traditional or electronic, is dancing the happy dance these days? Few that I know of. Most are about as cost-conscious as I have ever seen, as competition has everyone not only managing pennies but looking at cost in terms of fractions of a cent. So be careful what you wish for; the costs you save may be coming out of the execution quality and/or services that you desire.