"God helps those whom help themselves," he would say.

Despite my father's humble beginnings, selling magazines door to door, he was quite a teacher. He didn't finish college, but his outgoing personality and commitment to understanding his client's business, helped him become a successful salesman and entrepreneur. As a result, he never took anything for granted and had no patience for anyone needing spoon-feeding. Throughout my childhood, if I asked him a question, it was more than likely that he would point me to the bookshelf. "God helps those whom help themselves" he would say. Words of wisdom.

He often showed disdain for what he called the "little people" in our country that believe whatever they hear and refuse to actually research issues before forming opinions. He hated news media that sensationalized stories and anyone that pandered to the masses with promises or platitudes.

Listening to many commentators comparing the events of August 24th to the crash of 1987, prompted me to think about the events of that October. At that time, I was in my third year at Morgan Stanley in the IT department, and responsible for the firm's electronic equity trading systems, including program trading. During this period my father, was bedridden with Guillain-Barre syndrome and bored out of his mind, voraciously reading anything and everything he could get his hands on about the stock market and economy. One day, in late September, he informed me that he had just liquidated every cent of his stock market holdings. I said "why Dad?" He replied the "little people" were buying stocks in droves and brokers were trying to sell him "guaranteed" products, promising risk free stock market returns." He asked me to explain these" risk free" products to him, so I explained that portfolio insurance was based on the ability to sell futures quickly in case of a falling market. Being a quick study, he immediately understood the precariousness. Trade deficits were climbing and the dollar was weakening. He reasoned that meant that foreigners were getting antsy. I, of course, was way too smart to listen to him, but on that fateful day in October when the crash occurred, at least I knew that my father's brain was working very well, indeed.

I apologize for my brief walk down memory lane, but reading all the stories about August 24th and the uproar over the imperfections in the market that day, reminded me of 1987 and my father. To be clear, the 1987 crash was an epic failure in market structure. Many stocks had no buyers at all for long periods of time and the magnitude of the move was completely unprecedented. The system of human specialists and OTC market makers completely failed and it took weeks for the market to regain equilibrium. Whenever people make comparisons between modern events and that day, I still remember traders refusing to answer their phones for fear of having to make prices and how our systems were totally overwhelmed. I also remember my father, who despite being self-taught, knew enough to get out of the way of the herd...

So, when people talk about Rule 48, trading halts, and how the market performed for ETFs on August 24th, I flash back my father's words of wisdom:

• People always act in their own perceived self-interest and be very suspicious if someone tells you they don't
• Never oversimplify or accept simple explanations at face value; do the research yourself and gain understanding

Self-interest is a large issue in the debate over Rule 48 and the opening process from the 24th. Some have argued that it is inappropriate for DMMs to both control the opening print and be allowed to profit from trading at that price. But there are two sides to the story. While I agree it's wrong for any trader to have an incentive to set a price inappropriately, it's also true that the NYSE is the only market that guarantees their clients that market on open orders will always get executed at the opening price. Absent the ability for the DMM to correct that imbalance, this "feature" of the NYSE market would not work. Of course, the fact that the NYSE stopped disseminating price indications, using Rule 48, does make the potential conflict of interest worse. This is due to the fact that those price indications are used by other market makers to provide liquidity that might minimize price moves. When Rule 48 was invoked, however, only the DMM were aware of the magnitude that certain prices, such as KKR, were about to move. The only reasonable answer, in my opinion, is to also change the opening process on days where the exchange invokes rule 48, to be more automated.

I believe the NYSE should make a concerted effort to open stocks at or as close to 9:30 as possible, particularly on days with large market wide moves. Whether they do so by having the DMM start the opening process earlier, or automating the process on such days is up to them. It is, however, important to recognize that there is a large "public good" of the national equity exchanges being synchronized.

While I do not think that the NYSE should be forced to implement a specific method to synchronize their opening process with the other exchanges, we should not be ambivalent about the process they choose. On the contrary, one of the main lessons of August 24th should be that we need to improve the execution quality analysis of orders traded on all stock exchanges. In particular, all trades handled at the open and immediately afterward by all exchanges should be measured to see how well they meet their "best execution" obligations to the marketable orders they receive. That comparison, which is not currently part of Rule 605, would benchmark the opening and subsequent trades against the NBBO created by aggregating accessible quotations from open, competing venues. This would provide a mechanism for all market participants, including the regulators, to see major divergences, such as those that occurred on the 24th. This is the only way that I can think of, to manage the inherent conflicts of interest created by having traders with profit motives also in charge of setting prices.

In addition to the issues surrounding the open of NYSE stocks, the 24th exposed serious concerns about the market structure of trading popular Exchange Traded Funds (ETFs). Many self-interested parties have used this day to reprise the populist mantra that the "evil HFT bogeyman" is ruining trading. Others have used it to lobby for their favorite regulatory proposal to protect their businesses. Still others have used it to object to the growth of ETFs as an asset class. The fact is, there are issues with the current lack of coordination between stock listing markets and ETF listing markets, but a couple of changes would help restore order and improve the products.

Before going into the suggested changes, it's important to understand the core tenets of the ETF boom. In addition to being operationally more efficient and having lower fees, ETFs, under normal circumstances, can be traded intra-day, while mutual funds only trade at the close. At the same time, ETFs are designed to trade very close to their Net Asset Value (NAV). This is due to the ability of market makers to either create new units or redeem units into the underlying assets whenever the price moves too far from the NAV. Unfortunately, on the 24th, many ETFs diverged substantially from their NAV, prompting many to make statements about the inappropriateness of the equity market structure for ETFs. In my humble opinion, these comments are major over-simplifications.

The equity market structure works incredibly well for ETFs under normal circumstances. As I have previously noted the SPY (S&P ETF) actually trades with tighter spreads than the S&P future and has led to major cost savings for asset allocators. Other ETFs, including the QQQ and sector ETFs, such as the XLF Financial Sector IShares product have had similar impacts. That said, there are two issues that were exposed by the extremely volatile market on the 24th: First, trades in ETFs should be measured differently from equities, since the underlying issues can provide an objective benchmark and bid offer spreads may get out of sync with valuations. Second, ETFs should not be halted solely based upon its own price movements. Halts should also occur when a significant number of their constituent equities are halted or when they deviate too far from their NAV.

Improving upon the "best execution" reporting for ETFs, to properly handle volatile situations, would not be conceptually very hard. Firms such as ours could use the issuer-provided baskets and cash components to calculate an aggregate bid-side redemption price and an ask-side creation price per share of ETFS. In situations where there was a divergence between such NAVs and the execution prices provided by market makers, we could report on exceptions to routing firms. This ability could potentially save investors a lot of money. Armed with such exception reporting, routing firms could pressure market makers to improve upon prices that, in retrospect, look very bad. In addition, they could use the data to design more rigorous edits that could stop inappropriate orders from being sent in the first place. ETF issuers could also use such reporting to award coveted lead market making positions on the venues where those are available.

Another recommendation would be to rework ETF trading halts. For those ETFs whose underlying assets are normally traded during US business hours*, trading should halt whenever a significant number of their component securities are also halted or are yet unopen. In these situations, market makers find it difficult to price the value of the ETF, so bid/ask spreads can become extremely wide. ("Best execution" reporting would also require estimation) When wide spreads are present in the market, investors can lose a lot of money by thinking that they will receive a value approximating the actual NAV, while actually receiving substantially lower proceeds. My proposal would be for the ETF listing exchanges, along with their issuers, to set these thresholds, but that they should apply to all venues trading the products.

As my father used to tell me, if a job is worth doing, it is worth doing well. (Actually, he used to tell me not to do a job "half-assed"). The lessons of August 24th should drive us towards modernizing the opening process and at the same time, improving our measurement of those processes. It should lead us towards improving the market structure for ETFs, while also pushing us to improve reporting on the trading of those products. In a nutshell, the more we can measure, the more we can improve things.

*It is worth noting that there are several popular ETFs that trade based on overseas securities that are not normally traded during US hours such as the EWJ, Japan MSCI Index IShares. In those cases, market makers have sophisticated correlation models to provide reasonably tight bid offer spreads that are not used for the US products in general

David Weisberger, Managing Director, Trading Services at Markit
Posted 1 October 2015

S&P Global provides industry-leading data, software and technology platforms and managed services to tackle some of the most difficult challenges in financial markets. We help our customers better understand complicated markets, reduce risk, operate more efficiently and comply with financial regulation.

This article was published by S&P Global Market Intelligence and not by S&P Global Ratings, which is a separately managed division of S&P Global.