Best Mid-Sized Buy-Side Institution
Seattle is better-known for its coffee than for the sophistication of its financial institutions, but when it comes to navigating the markets, Rainier Investment Management is no less savvy than its counterparts in New York and London.
“We believe that an understanding of market structure and attention to detail in regards to broker and exchange order-routing practices can make the difference between realizing low fill rates combined with an excessively high level of opportunity costs, and high fill rates combined with lower than average opportunity costs/market impact,” said Justin Kane, director of equity trading at Rainier.
Rainier manages about $13 billion, mostly in U.S. stocks across large, middle, and small capitalizations. The company employs eight managers for domestic portfolios and three who focus on international holdings.
Portfolio turnover is about 125%, implying Rainier buys and sells about $16 billion worth of stocks per year; smaller-cap names trade a bit more frequently than larger companies. Rainier moves everything through the same order management system and execution management system.
Inforeach is Rainier’s outsourced EMS. “We use a hosted version—that allows us to leverage the tech team at our EMS vendor, an enormous benefit considering our smaller size,” Kane told Markets Media. “This eliminates many obstacles that we would otherwise face as a small money manager with limited IT resources and specialized technical expertise. Larger money managers may have IT staffs that are half the size of our firm.”
Along with trading-desk colleagues Tyler Platte and Matthew Steadman, Kane has moved proactively to reduce ‘trading friction’—market-speak for poor executions that boost costs and reduce the end-user investor’s net return—in Rainier’s portfolios. The adversary can be high-frequency traders and other market players with ultra-short time horizons of seconds or less.
“Today’s marketplace is littered with HFT ‘scalping’ strategies dependent on the scores of complex order types utilized to efficiently assess supply and demand changes occurring in the marketplace at any given moment,” Kane said. “As a result, Rainier has taken several steps that advance the way we source liquidity and interact with counterparties.”
With regard to the use of brokers’ algorithmic trading suites, Rainier’s efforts include configuring and customizing algos, including allowing for the avoidance of certain trading venues; requiring that all algo suites be directly owned by the sponsoring broker to ensure the highest level of control over order-routing practices; avoiding the algo suites offered by large wholesale market-making firms; and minimizing the use of smart order routers, instead opting to utilize sponsored direct market access when appropriate.
Regarding DMA, Kane said Rainier will not accept a broker’s algo trading suite without direct access to a minimum of five major exchanges. In addition, Rainier requires access to a standard set of order types and customized order routing at each exchange, allowing for more control over order types versus solely relying on algorithms and smart order routers; according to Kane, this functionality has enabled Rainier to verify whether certain order types were being properly managed at the exchange level.
“We believe that implicit costs can be lowered substantially if we partner with brokers who prioritize fill rates and execution quality ahead of managing exchange access fees,” Kane said. “DMA substantially increases our level of control when it comes to order routing, not only at the broker level, but at the exchange level.”
“DMA allows us to use the various order types native to each exchange,” Kane continued. “Some of these order types can be defaulted to be non-routable or configured to immediately route out. Many folks don’t realize that if the exchange is sent an executable order that it can’t match up with their own order flow, they too may be required to pay the ‘take’ fee at another exchange.”
A nine-year Rainier veteran, Kane held previous buy-side trading positions at Westpeak Global Advisors and Safeco Asset Management, and he also worked in sell-side trading at Sherwood Securities/NDB Capital Markets, which was acquired by Deutsche Bank in the early 2000s.
Rainier traders actively measure their performance via trade cost analysis. “We have both real-time TCA that’s able to benchmark against various strike prices, and calculate interval (volume-weighted average price) on the order and individual broker-placement levels,” Kane said. “The order level will aggregate multiple broker placements, irrespective of whether or not an order had been executed using different venues.”
A Rainier trader may use ITG’s Posit or Liquidnet to execute an order, the execution of which Rainier’s TCA evaluates while taking into consideration factors such as order size, urgency, and market volume. “Since ITG has added a post-trade cost model that takes into consideration realized volume rather than historical volume profiles, we have found that the analysis has become much more meaningful,” Kane said. “After several years of work we now believe we have finally have the ability to produce actionable TCA in order to help us better assess both broker performance for each placement in aggregate, and trader performance at the order level.”
“Today’s marketplace is littered with HFT ‘scalping’ strategies…As a result, Rainier has taken several steps that advance the way we source liquidity and interact with counterparties.” Justin Kane, director of equity trading at Rainier
‘We can compare how successful we were in mitigating slippage, and we’re able to collect a fair amount of information that increases transparency beyond what your traditional small money manager is able to obtain,” Kane added.
Venue analysis included in Rainier’s TCA showed that in the last nine months of 2012, the firm used electronic block crossing networks to trade 17% of its shares, and that venue type beat other venues by 20 basis points to 30 basis points on average. “It was the amount of outperformance that was surprising,” Kane said. Rainier’s analysis has also showed its cash desk trading more efficiently than its algorithms, on average. “When we dug down we determined that the outperformance could be attributed to the traditional cash desk sales trading coverage that had been able to successfully negotiate block trades,” Kane said. “We believe the ‘block negotiation skillset’ combined with a firm’s institutional order flow had enabled several firms to substantially outperform other sales traders on the cash desks who simply ‘worked’ most or all of their order in the market.”
“It appears the outperformance reflected in implicit costs is very likely to be significant enough to more than make up for the additional commission savings that our client would otherwise realize in explicit costs when ‘working’ orders in lower-cost venues such as algorithmic trading suites,” Kane continued. “This should come as no surprise since block crossing, which may include price improvement at the mid-point, helps to mitigate both market impact and opportunity cost,” or the cost of trading too slowly or not trading at all.
When Rainier does trade with algorithms, the firm seeks maximum transparency into broker order routing via Financial Exchange Information (FIX) technology. “Rainer has required all broker-dealers to provide us with the ability to route orders electronically via a direct FIX connection and support real-time TCA capabilities,” Kane said.
Rainier conducts due diligence on brokers that covers the configuration of dark pools and order-routing practices. As conducting due diligence on many broker-dealers is time-consuming, the firm has leveraged Commission Sharing Arrangements to reduce the broker-dealers it needs to oversee by about 30% since 2011.
One 2013 initiative for Rainier’s trading team is drilling down further in analyzing execution quality by order size. Rainier technologists recently finished programming the firm’s EMS to automatically tag any execution over 10,000 shares. “We are now able to segregate block crossings from working orders,” Kane said. “We are eagerly awaiting our first full-quarters’ worth of results from ITG.”
As a $13 billion manager, Rainier’s trading-technology budget can’t compete with institutions that are 10 and 20 times its size, so it has to strategize for how it can effectively fight above its weight class. One example of this is how Rainier implemented broker neutrality on its EMS back in 2009.
“Setting up processes that are scalable is key,” Kane said. “You want it to be scalable and not spend a significant amount of resources. And if you can automate, the output will be more accurate, actionable and valuable.”