Capital Markets Broker Dealers

Transcription

1 Industry Update Joel Jeffrey Niamh Alexander Brian Kleinhanzl Brian Gardner Michael Needham, CFA Kyle Voigt Capital Markets Broker Dealers April 15, 2014 Summary-- This report focuses on the potential fallout on capital markets and market structure stocks from any changes in the current regulatory environment following the increased focus on high frequency trading (HFT) and payment for order flow (PFOF) as a result of the publication of Michael Lewis s book Flash Boys. Stocks in our coverage universe that are associated with these business practices have declined by an average 7% since the market close on March 28 th versus a 1% decline in the S&P 500. While we believe that both practices are likely to come under heightened scrutiny by regulators, we believe that any potential changes will involve a relatively substantial overhaul to market structure regulations, making this a longer-term event. Key Points-- Potential Regulatory Change: We believe the most probable near-term action against HFT would come from the state Attorney General (AG) offices, most likely the NY AG, and would be targeted at specific market participants rather than the broader industry. Congress will likely raise the issues surrounding HFT in upcoming hearings but is unlikely to directly intervene, in our view. We expect the SEC to act deliberately and methodically in reviewing Reg. NMS and HFT. E-brokers Impacted by Scrutiny over PFOF: Interestingly the e-broker stocks, which are more impacted by the risk to PFOF than HFT, are down 10% on average despite the fact that PFOF was referenced only a handful of times in the book. The market structure stocks, which are more directly linked to HFT, are down just 7% on average. While these two issues are tied together through the current market structure in the U.S., we argue that they are two separate issues. AMTD and ETFC would be the most impacted by any restrictions on PFOF, in our view. Headline Risk Could Weigh on Exchanges & Specialty Brokers: We believe NDAQ and CME are the public exchanges that could be most negatively impacted by a reduction in HFT. KCG also operates HFT desks in several markets, but could be more negatively impacted by rules aimed to limit off-exchange trading, in our view. While we think the path to regulatory change could be long, headline risk could continue to weigh on the group. Traditional Market Makers Could Benefit from Reduced HFT: Longer term, traditional broker dealer market makers such as IBKR could benefit if high frequency trading declines, as we believe this would widen bid ask spreads and return volatility to more normalized levels. We Cover the Following Directly and Indirectly Impacted Sectors in this Note: (1) Securities Regulation; (2) E-Brokers; (3) Exchanges/Market Structure; and (4) Universal Banks. Keefe, Bruyette & Woods, Inc. does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their investment decision. Please refer to important disclosures and analyst certification information on pages

2 Regulatory Outlook There are three main participants as we look at the potential political reaction to Michael Lewis s book, Flash Boys the state AG, the federal securities regulators, and Congress. We think the state AGs have the potential to be the most disruptive especially in the case of the NY AG since they are often less constrained than federal regulators and are generally less predictable (in our view, criminal and civil investigations are tougher to determine since they generally occur outside of the public sphere). We expect the SEC to be measured and cautious in whatever action it takes. Finally, we expect Congress may weigh in with hearings, but we expect no legislative action. Michael Lewis's Flash Boys has intensified the attention being paid to high frequency trading (HFT) and we expect that upcoming policy events will continue to keep HFT in the headlines. However, we do not think that the book will lead Washington to make quick changes in rules governing HFT largely because of the complexity of trading rules and the fear among regulators of unintended consequences. There may be many at the Securities and Exchange Commission (SEC) who recognize that electronic trading has benefited investors, including retail investors, with lower costs for trading and greater liquidity so the SEC would be reluctant, in our view, to move quickly with proposals that could undo these benefits. The most likely action from the SEC is, in our view, pilot programs and increased disclosure requirements which would not unduly disrupt the markets. The SEC has been examining market structure issues since 2005 when it finalized Rule NMS. In 2010, the SEC proposed changes to the trade-at rule which would increase transparency and disclosure for some trades in dark pools and subject dark pools to display best-priced offers. The Financial Industry Regulatory Authority (FINRA) has also proposed increased disclosure requirements for trades occurring in dark pools. Other market structure initiatives undertaken by the SEC include the 2010 Concept Release, an SEC website dedicated to market structure data and analytics, and a pilot program to increase tick sizes for certain small company stocks. We believe that SEC chairman Mary Jo White wants to improve the SEC's image and does not want to see the agency trumped by other regulators or to be seen as being behind the curve. However, she has also demonstrated that she is deliberate and methodical and will not be rushed into action just in order to improve her agency's political standing. In recent public appearances (a congressional hearing and a meeting of the SEC s Investor Advisory Committee), Chairman White indicated that the agency is aware of issues in electronic trading and that the issues raised in the Lewis book are not new. She further indicated that the SEC has several investigations going on related to HFT. As for new regulations, she said that the agency will be thorough in its examination of the markets and would be data driven. We should note the SEC needs, in our view, to be deliberate in crafting changes to Reg NMS and any related rules since the SEC needs to abide by cost-benefit analysis guidelines. Failure to do so in the past has resulted in some SEC rules being overturned in court. We believe that Chairman White wants to avoid a repeat of such an outcome. The SEC s most likely action, in our view, would be to proceed with pilot programs for increased tick sizes for smaller companies with potentially a trade-at rule, as well as finalizing disclosure rules for dark pools. In our view, the unknown(s) are the state AG. New York AG Eric Schneiderman has expressed interest in looking at HFT and what he calls "insider trading 2.0." In recent speeches, he has mentioned co-location as something he is closely monitoring and has endorsed the concept of "frequent batch auctions." It is not clear to us how he would implement the latter given his role as a law enforcement official and not as a broader market regulator. That being said, the NY AG is clearly investigating HFT Please refer to important disclosures and analyst certification information on pages

3 and his office has demonstrated a willingness to pursue legal actions ahead of the SEC. We should note that Mr. Schneiderman has publicly said that he views the characterization of the markets as rigged to be hyperbolic so any investigations by his office may focus on specific industry participants and not industry-wide, in our view. Also, the Massachusetts Secretary of State William Galvin recently sent surveys to investment advisers asking about their use of HFT. At this time, no congressional hearings are scheduled on the HFT issue, but we expect hearings on this or related issues in the coming months. We expect that SEC Chairman White will likely testify again in Congress as part of annual SEC oversight hearings and the HFT issue will likely be raised at these hearings. We doubt Congress would pass legislation addressing HFT this year because 1) we doubt much legislation will pass Congress this year due to the election year environment, and 2) because this is a technical and complex issue that Congress, in our view, would prefer to leave to the regulators. Legislation aimed at the HFT industry might be introduced but chances of any bill moving far in the legislative process are remote, in our view. The Lewis book again raises the prospect of a financial transaction tax but other than increased headlines from the proposals' sponsors, we see no signs that Congress's appetite for passing such a tax has increased. We continue to believe that chances of a financial transactions tax passing Congress are quite low. The bottom line is that we expect that the SEC will act deliberately and methodically in reviewing Reg NMS and HFT. Congress will likely raise the issues in upcoming hearings but is unlikely to directly intervene, in our view. Outside of Washington, we expect that state officials, especially the NY AG, will continue to investigate HFT and may bring enforcement actions against some entities if the NY AG believe it can prove that fraud has been committed, but we do not believe that the NY AG will impose market-wide changes to the industry. So far the NY AG has fined some market data companies for releasing data early to those willing to pay for such access and ended that practice, but there have been no actions against the professional trading firms themselves or anything against exchanges, other than a call to end co-location but there s no regulatory reason to end it. E-Brokers: Payment for Order Flow Has Been & Likely Will Continue to Be a Strong Revenue Source Although payment for order flow was mentioned only a handful of times in Michael Lewis s book, the prospect of its elimination resulted in a meaningful sell-off in shares of AMTD, ETFC, and SCHW. Payment for order flow has been a reliable and profitable source of revenue for the e-brokers and has from time to time drawn the attention of market regulators. The three large publicly traded e-brokers send client trade orders to external brokerage firms to execute orders and receive a fee based on the number of shares or contracts. Although a customer may request that their order be executed on a specific exchange or trading venue (directed), almost all orders placed through the e-brokers in our coverage universe do not designate a specific execution venue (non-directed). It is these non-directed orders that allow the brokers to route orders to specific trade execution venues in return for per share or per contract fees. Given this flexibility and within the context of getting the client "best execution," brokers typically chose destinations (market makers or execution venues) that paid them the highest amounts. Despite the perception that customer orders are being sold and that retail customers may not be receiving the best execution, the practice of payment for order flow provides some benefits to the retail investor. In fact, the market orders that are routed to a market maker typically receive some level of Please refer to important disclosures and analyst certification information on pages

4 price improvement, although usually quite small. In exchange for routing the order flow to the market maker, the e-broker receives a per-share fee. Why does the market maker pay for retail order flow? Primarily because retail clients tend to be longer-term holders of stocks and are not likely to quickly move positions; the market makers are able to use this order flow to trade with other larger market participants. Without the consistent and longer-term order flow, market makers would have to increase their risk levels which could result in wider spreads. This could make trading more expensive for retail investors as market makers look to be compensated for their increased risk exposure. While much of the recent attention on PFOF has been focused on the equities markets, we would note that it is also common in the options markets. In fact the per-contract fees paid for trading in the options markets are significantly higher than the per-share fees paid in the equities markets. Given the surge in options trades as a percentage of total trades at the e-brokers in the past five years, we estimate that PFOF for options trades now accounts for roughly half of their total PFOF. To illustrate the impact that options trades have on PFOF, we would highlight AMTD, which generates the largest percentage of DARTs from options trades. Management disclosed that options accounted for 34% of total DARTs in F1Q14, or 141K trades per day. Based on the firm s December quarter-end Rule 606 report, we estimate the average payment per options contract that AMTD received was $0.46. However, AMTD and the other firms do not disclose the average number of contracts per trade, so accurately estimating the amount of revenue generated by options trades is difficult. As a proxy for AMTD, we would note that IBKR disclosed that in 3Q13 its average number of contracts per trade was nine. Using this number of contracts and AMTD s average PFOF per contract, we estimate that options contribute more than 50% of PFOF revenue. While we believe that regulators are unlikely to make any changes to PFOF in the equities market, it is even less likely that they will curb this practice in the options markets given the lower levels of liquidity and more professional nature of this market. Does Payment for Order Flow Result in Best Execution? While there is no specific definition for best execution from the SEC, the website states that Brokers are legally required to seek the best execution reasonably available for their customers' orders. To comply with this requirement, brokers evaluate the orders they receive from all customers in the aggregate and periodically assess which competing markets, market makers, or electronic communications networks (ECNs) offer the most favorable terms of execution. Some of the factors a broker must consider when seeking best execution of customers' orders include the opportunity to get a better price than what is currently quoted, the speed of execution, and the likelihood that the trade will be executed. Based on the preceding paragraph, we believe that PFOF is not inconsistent with the SEC s definition of best execution. Specifically, the modest price improvement and the immediate execution of trades placed with a market maker appear to conform to the factors that the agency highlights for consideration when seeking best execution. That said, one could argue that the modest levels of price improvement (often sub one penny) on the order offered by the market maker is not enough to justify the payment that the broker is receiving, and that a better execution of the order may have been available in a different venue. Quantifying Payment for Order Flow for the E-Brokers Fees from PFOF differ by security class and broker-dealer, and based on company disclosures ETFC and SCHW have agreements with external broker-dealers to route them a majority of client order flow. SCHW has had an agreement with UBS for a number of years, and in the past UBS has received more than 90% of SCHW's client order flow. ETFC has an agreement with Susquehanna International Group Please refer to important disclosures and analyst certification information on pages

5 who, in October, 2013, purchased ETFC's market maker G1 Execution Services for $75 mil. which included an agreement to receive at least 70% of client orders from ETFC for five years, according to ETFC filings. According to the latest SEC 606 report, AMTD routed approximately 40% of client orders to Direct Edge, 30% to Citadel, and 20% to Citigroup. ETFC is the only firm that discloses payment for order flow revenues in its financials, but SCHW has also disclosed its order routing agreement with UBS. AMTD is the most difficult of the three to estimate so our analysis assumes the company earns approximately the same PFOF as a percentage of trading revenues as ETFC. Exhibit 1: Payment for Order Flow Estimates 2013 Disclosed PFOF (MM$) % of Trading Revenue Assumptions Estimated 2013 PFOF (MM$) 1 Estimated 2013 PFOF % Rev. Estimated 2013 PFOF % EPS Where is PFOF Reported in Financials? ETFC % Disclosed PFOF % 21% Fees and Service Charges AMTD NA NA PFOF = 20.5% of Trading Revenue % 18% Commissions and Transaction Fees SCHW % PFOF = 8.2% of Trading Revenue % 5% Other Revenue (1) ETFC PFOF estimate increased by $13.5 mil. to reflect estimated order flow sent to ETFC's market maker G1 Execution Services Source: Company reports and KBW Research As expected, payment for order flow accounts for a high percentage of earnings at ETFC and AMTD relative to SCHW where commissions account for only 16% of revenues (vs. 42% at AMTD and 24% at ETFC). That said, we would highlight that our estimate for SCHW s PFOF revenue is meaningfully lower than our estimates for AMTD and ETFC relative to their trading revenues. Based on SCHW s Rule 606 disclosure, the firm earns, on average, less than $ per share for orders executed through UBS, well below ETFC s and AMTD s per-share average for their orders. Although we estimate SCHW earns around $75 mil. of PFOF per year, the firm s commission revenues are nearly double ETFC and within 25% of AMTD. We expect that when SCHW s order flow agreement with UBS expires in 2015 there could be an opportunity for SCHW to negotiate a more favorable order flow agreement, assuming there is no change to the current regulatory structure. KCG is one of the five large wholesalers of retail liquidity by paying for order flow (we estimate KCG pays $ million in PFOF annually), and it sees significant order flow from the retail brokers. Even if PFOF were to be eventually eliminated by the regulators (not even a consideration right now), we still think the wholesalers would have a significant advantage over other brokers for garnering retail liquidity especially because of their established connections to the thousands of retail brokers, the processes and procedures in place to provide the retail brokers with their required regulatory standards and reports, and their liquidity advantage. In options markets, IBKR's market maker has substantially reduced its payment for order flow and barely does it at all now. We therefore don't see much changing in that side of the business; however, if HFT is reduced in options then this could drive back up IBKR's margins because we'd expect bid-ask spreads to widen as a result. We note that IBKR's brokerage segment does receive revenue from PFOF, mostly in the form of rebates from options exchanges. We estimate IBKR's PFOF could be up to 5% of its earnings. Although PFOF currently represents approximately 20% of earnings for ETFC and AMTD, we estimate that the bottom line contribution will decline as the firms approach their normalized earnings potential largely due to the offset from higher net interest revenue as interest rates rise. Although we believe it is more likely than not that the e-brokers will continue to get paid for PFOF, we calculate our normalized EPS forecasts excluding PFOF in Exhibit 2 below. Potential upside in Exhibit 2 represents the difference between the firms' three year average P/E ratios and their current price-to-discounted normalized EPS ratios. Removing PFOF from our forecasts reduces the projected upside by Please refer to important disclosures and analyst certification information on pages

6 approximately 10-15% for the e-brokers. Given the roughly 10% decline in the e-brokers, we believe that the potential loss of PFOF is largely priced into the stocks. Exhibit 2: Normalized EPS Forecasts With & Without PFOF Normalized EPS Forecasts P/E Multiples Upside Potential to Current Share Price Based on Normalized EPS Forecast Upside Potential to Current Share Price Based on Normalized EPS Forecast (Ex. PFOF) Normalized EPS (Ex. PFOF) 1 P/E: Disc. Norm. (Ex. PFOF) 1,3 P/E: 3Yr Avg 2 Upside (+200bp Rates) Upside (+100bp Rates) Upside (+200bp Rates) Price Normalized EPS 1 P/E NTM P/E: Disc. Norm. 1,3 ETFC % -4% -8% -19% AMTD % 10% 21% 0% SCHW % 14% 35% 11% (1) Assumes 200 bp increase in interest rates and that normalized earnings is reached by 2016 for ETFC & SCHW and 2019 for AMTD. (2) ETFC 3-yr average is not relevant in our view. 16.0X is the KBW normalized estimate. (3) P/E ratios represent the current share price divided by our forecasts for normalized EPS (discounted back to year-end 2014). Upside (+100bp Rates) Source: Company reports and KBW Research It is our view that interest rate expectations will dominate stock performance going forward, and based on our normalized EPS forecasts we continue to think that AMTD and SCHW have meaningful upside potential in a higher interest rate environment, even if payment for order flow were eliminated. Moreover, we believe that our normalized EPS estimates excluding PFOF are conservative due to 1) our belief that PFOF is more likely than not to remain in place, and 2) we assume that PFOF is not supplanted by other measures such as internalizing customer trading at the e-brokers. Market Structure The concerns raised in Flash Boys regarding the fairness of the current market structure could eventually lead to rule changes that limit HFT or changes in customer behavior that could impact market structure companies. We believe that the stocks most likely to be impacted include KCG, ITG, IBKR, NDAQ, and ICE. Additionally, CME could be negatively impacted if there are changes to the futures markets as well. We believe eliminating HFT firms, or greatly reducing their numbers, would not be devastating to the business models of the securities exchanges as their revenue sources are diversified. However, the impact of this change would not be inconsequential on the exchanges, particularly CME and NDAQ, which have a greater relative proportion of revenue from this customer group. Additionally, even if HFT is not a big driver of revenue, the removal of this business stream could have the tertiary effect on a firm like NDAQ by lowering demand for some services that other customers use in order to have similar access that the HFT community enjoys. Besides the potential impact of specific changes to the current market structure, we believe it will take some time to make meaningful changes but that potential pilot programs and headline risks could impact the stocks in near future. We believe that it will likely take years for any meaningful changes to be implemented as the regulatory bodies that oversee the securities and futures markets have historically been very deliberate in implementing change. For example, in 2010 the SEC issued a concept release to consider changes post Reg NMS implementation. However, none of the broad changes considered in that proposal have been implemented although the process of passing Dodd-Frank used up a significant amount of the SEC's resources resulting in the pushback of market structure reform deadlines. In the futures market, the CFTC issued a concept release last year that hasn't resulted in any rule proposals yet or any changed behavior. This is not to imply that no action will be taken by the regulators. In fact, the SEC could decide to launch some pilot rule changes that have come up in previous concept releases that could push some volume back to exchanges and limit dark execution (including a Trade-At pilot). That said, while meaningful changes to existing market Please refer to important disclosures and analyst certification information on pages

7 structure will take time to be implemented, the headline risk associated with concerns over HFT will likely remain as a potential headwind for the market structure stocks, in our view. With High Frequency Trading an Estimated 50% of U.S. Equity Volume, We Believe Its Best Days Are Done After 2009 Volatility Spikes and Velocity of the Liquid Stocks Peaked One indicator of the presence of high frequency trading in the market is the velocity of the US equities, especially in the most liquid stocks where HFT firms participate most. In the charts below we see how velocity of the most liquid stocks 1) peaked in a volatile 2009, and 2) is at least 50% lower in most stocks, except in the small-cap stocks where HFT had still been growing. Recall Tape B is the old Amex tape and primarily ETFs where statistical arbitrage strategies are commonly deployed. We believe the already declining presence of HFT is due to lower volatility, the shrinking presence of core fundamental investor trading volume which HFT most profitably trade against, some rules limiting behavior and increased surveillance of market behaviors by this customer group. Exhibit 3: Velocity in Top Volume Stocks Has Fallen Since 2009 Velocity of Top 25 Volume Stocks Tape A Tape B Tape C Russell 2000 NYSE Listed Nasdaq Listed X 53.6X 4.2X 7.5X X 58.1X 5.5X 8.7X X 80.7X 7.1X 10.0X X 76.2X 8.1X 10.4X X 152.6X 10.1X 9.7X X 132.3X 9.7X 9.2X X 64.2X 8.4X 8.9X X 39.4X 4.5X 6.9X Source: FactSet, NYSE Arcavision, KBW Research. Please refer to important disclosures and analyst certification information on pages

8 Exhibit 4: High Frequency Trading Represents Just Less Than 50% of U.S. Equity Volume, Down from 61% in 2009 Average Daily Volume (billions of shares) U.S. Equity Market ADV Other HFT 61% 52% 56% 55% 51% 49% 21% 26% 35% Source: Tabb Group, WFE, and KBW Research Who Loses/Wins If HFT Is Curtailed or Substantially Reduced, PFOF Changes Wholesale Market Structure Exhibit 5: Estimated Range EPS Impact if HFT Market Participation Is Significantly Limited or Reduced Over Time Estimated EPS Impact Ranges CME 20-40% NDAQ 5-20% ICE 5-15% CBOE 1-10% KCG 1-30% Source: KBW Research Exchanges: CME, NDAQ, ICE, CBOE Exchanges generally don't separate out the proportion of revenue coming from the High Frequency Trading group for any specific revenue line or as a total level in aggregate for the firm. However, exchanges have HFT firms as customers paying fees in several areas including co-location fees, access fees, trading fees (though in equities the HFT community is more likely to be collecting fees for providing liquidity than paying trading fees) and proprietary market data fees for more depth of book pricing than the standard market tape data. Of the four exchanges in the U.S., we believe that CME has the biggest portion of revenue and EPS coming from HFTs as CME management estimated that the group represents less than one-third of total revenue (in 2012). Despite the apparent risk for CME from changes to rules regarding HFT, it's important to note that there is no apparent movement, concept releases, or regulatory pressure to limit this HFT in the futures markets at this time. The CFTC had formed a committee to help better understand, define, and establish ways to best monitor HFT, but there are no proposals in place to change anything. The book focuses more on the securities markets for equities and options and the maker-taker pricing structure, which does not exist in the futures markets. We believe NDAQ may also have exposure to HFT customers in terms of EPS though revenue sources Please refer to important disclosures and analyst certification information on pages

9 such as access fees, co-location fees, and proprietary market data fees. In terms of disclosure, in 2012, NDAQ stated that it earned a net $50 million in revenue from this customer group globally, or 3% of total revenue at that time. More recently, at NDAQ's 2014 Investor Day, management stated that it earned $12 million of revenue from HFT in its U.S. Equities division with the group, up from $10 million in We believe that these data points understate the impact of a change HFT on NDAQ s business. For example, the reduction in demand from HFT firms for co-location or market data services could actually reduce demand from other customer segments, in our opinion, as this service is utilized by other customers in part to defend themselves against predatory HFT strategies. A significant curtailment of NDAQ's co-location services and reduction in demand for its proprietary market data services without a big change in its maker-taker pricing structure could affect 5% to 20% of EPS, in our opinion. Lastly, if NDAQ experiences market share loss due to losing HFT customers, it could also reduce NDAQ's U.S. share tape revenue, which is earned based on market share of trading. Therefore, declining HFT could mean lower customer demand in other areas for some of NDAQ's premium services where customers are paying for the proprietary products in order to remain competitive with the HFT firms in the industry. Less likely to be impacted by changes that could limit HFT are CBOE and ICE given their lower percentage of estimated revenue generated by HFT as these customers are less active in U.S. Options and ICE's European Futures markets. Besides the lower contribution to earnings from HFT than CME or NDAQ, we believe that ICE s recent actions indicate less focus on the HFT customer segment. Specifically, ICE's CEO, Jeff Sprecher (owner of NYSE & Arca) publicly criticized U.S. equity market structure and suggested eliminating or revisiting maker-taker pricing. In Flash Boys, the founders of IEX also state that ICE attempted to buy IEX, where the key concept is to prevent the speed advantage that HFTs have in executing orders. CBOE s biggest customer group is more the market makers and online brokers but it earns some revenue, especially on its futures exchange from HFT we expect HFT is also active in its more liquid options. Specialty Brokers / Market Makers: KCG, ITG, IBKR KCG has its own high frequency trading desks that operate in several markets including equities, options, futures, cash fixed income and several overseas liquid markets (part of the former Getco business). KCG's brokerage and market making operation also runs dark pools that may be collecting fees from HFT for access or facilitating better internalization rates for its own customer orders. We don t know if it is collecting such fees or if it gets better match rates on its own capital allocation because of how it interacts with customer orders which were some of the practices flagged in the book but not attributed to any specific broker. Its biggest business, the wholesale market maker for retail order flow, competes with high frequency trading firms but also internalizes some of the orders, and we think one potential fallout of market structure change could be some rules to push more orders and matches back onto exchanges at the expense of internalization. Please refer to important disclosures and analyst certification information on pages

10 Exhibit 6: KCG's Two ATSs Comprise Roughly 9.4% of All Dark Pool Trading YoY ADV Growth in 2013 (%) YoY Growth in ADV from KCG Algos and EMS vs. Consolidated Volume 20% 15% 10% 5% 0% -5% -10% -4.2% Consolidated U.S. Equity ADV +18.8% KCG Algo & EMS ADV 2013 ADV (millions of shares) KCG ADV and Market Share of Dark Pool Volume % % GETMatched (ATS) Knight Match (ATS) Source: Company presentations and KBW Research ITG indirectly benefits from the presence of HFT in the market. Customers are willing pay higher commissions for its sophisticated algorithms that are geared to be anti gaming to protect them from HFT when executing in ITG s venues. ITG also engages in riskless principal matching of customer orders that might slow in the event of more limitations on internalization. However, we'd expect ITG to be more of a beneficiary than a loser if several of the competitive faster dark pools, which offer low execution size, were eliminated or shuttered. This could result in more market share to ITG s longstanding wholesale dark pool where large blocks are matched. IBKR would benefit from the reduction of high frequency trading in the options market as this would result in wider bid-ask spreads benefiting its market making business, in our opinion. IBKR has experienced substantial decline from its 60% operating margins because of competition from HFT, as illustrated in Exhibit 7. Reducing this customer group would likely be good for sophisticated market makers such as IBKR whose smart order routers and sophisticated technology are traditionally an advantage in that market. Exhibit 7: Penny Quoting That Accelerated HFT Participation in Options Hurt the Dominant Market Makers' Profitability IBKR Trading Gains (excl. FX gain/loss) $1,400 $3.00 $1,200 $1,000 $800 $600 $400 $200 $0 $2.50 $2.00 $1.50 $1.00 $0.50 $ E 2015E Trading Gains ($ Millions) Trading Gain per Options Contract Traded (Market Making) Source: Company reports and KBW Research Please refer to important disclosures and analyst certification information on pages

11 Potential Industry Change Ahead of Regulatory Change Despite the long lead time expected for any type of meaningful regulatory changes to the current market structure, it is possible that institutional and retail brokers could force change on the markets by altering their routing practices. While it is possible, we have seen very little movement to date by customers to force this change. For example, according to Lewis's book, it appears that IEX has been educating customers for some time now about some questionable practices in the market without much, if any, change to how institutional traders are directing their orders. IEX is matching less than 1% of the U.S. equity market volume and we're not yet seeing any significant change in the broker market shares. In the following paragraphs we highlight two other possible market driven changes: closing of fast dark pools and the end of co-location. Exhibit 8: IEX Continues to Gain Volume & Market Share IEX ADV IEX as % of total market volume ADV (millions of shares) % 0.5% 0.4% 0.3% 0.2% 0.1% 0.0% % of Overall U.S. Cash Equities ADV Source: IEX, BATS Global Markets, and KBW Research. April-TD as of 4/9/14. Voluntary Shuttering of the Fast Dark Pools at Broker Dealers - Possible, in Our Opinion Reg NMS brought about the proliferation of many of these fast 'dark pools' or alternative trading venues that essentially ended up facilitating matches away from the exchange but don't appear to have facilitated any bigger trade size with average trade sizes similar to or even below the exchanges (less than 300 shares) in these pools. According to the Lewis book, many of these pools are programmed to benefit the broker owner of the dark pool rather than act as a broker neutral alternative trading venue. In fact, based on a WSJ article published April 8th, Goldman Sachs is considering shuttering one of the biggest of these pools in the industry, Sigma X (75 million ADV in February-14, according to Tabb Group). Others dark pools include Credit Suisse's Crossfinder, Barclays LX, as well as KCG's two pools (Knight Match & GETMatched). KCG runs two dark pools and could see some liquidity decline as a result of customer pressure. In contrast, ITG runs one of the more traditional large block matching venues (20,000 shares plus) that could benefit if wholesale (large institutional orders) still seek alternative ways to find matches without exposing orders on exchange. We don't see any risk to ITG's POSIT or competitor Liquidnet's dark pool. If anything, they could benefit both of these pools are already set up with anti-gaming technology to prevent HFT accessing their institutional customers' liquidity. Voluntary End of Co-Location - Unlikely, in Our Opinion Co-location is the practice of moving servers as close to the market centers' data centers as possible in Please refer to important disclosures and analyst certification information on pages

12 order to gain a faster execution. HFT are willing to pay for this service and given our belief that the U.S. exchanges are generating a few hundred million dollars of revenue from co-location, we think it's highly unlikely that they end this service without being forced by regulatory change, which we believe could take a long time to implement. Exhibit 9: NDAQ Access Services and U.S. Proprietary Data Revenue Has Grown Significantly Since 2007 Access Services Revenue ($ Millions) $300 $250 $200 $150 $100 $50 $0 $75 $103 NDAQ Access Services Revenue (Includes Co-location Revenue) $139 $172 $224 $239 $235 $236 $ E 2015E U.S. Proprietary Market Data Revenue ($ Millions) $200 $180 $160 $140 $120 $100 $80 $60 $40 $20 $0 $88 NDAQ U.S. Proprietary Market Data Products Estimated Revenue $105 $117 $127 $134 $149 $151 $164 $ * 2014E* 2015E* *Because of a change in NDAQ's reporting format, we estimate 2013 and forward U.S. Proprietary data revenue by holding shared tape revenue from 2012 constant in those years. Source: Company filings and KBW Research. Regulatory Concepts Previously Floated If Reconsidered, Piloted or Put into Effect Could Limit HFT, Co-Location, Internalization Reducing the Fragmentation of Liquidity to ATSs/Dark Pools: In 2009, the SEC proposed lowering the volume threshold at which an ATS would be required to publicly display its best priced orders from 5% to 0.25% of trading in a given stock (though there could be exemptions for truly dark venues or large blocks). Additionally, in the SEC's 2010 concept release, the SEC floated the idea of lowering the volume threshold for which ATSs must provide fair access to their venues from the 5% volume level. This could potentially drive some volume back to lit venues, and actually benefit exchanges if they see higher fees from a greater market share. KCG runs two such dark pools and could be negatively impacted with seeing lower volume. Please refer to important disclosures and analyst certification information on pages

13 Exhibit 10: TRF Market Share Has Grown Significantly Since Reg NMS - Mostly Because of the Growth in These Faster Dark Pools Now an Estimated 10-15% of U.S. Equity Volume U.S. Equity Market TRF Volume Growth Average Daily Volume (billions of shares) On-Exchange ADV TRF ADV TRF Market Share (%) 60% 50% 40% 30% 20% 10% 0% TRF Market Share (%) Source: Arcavision, BATS Global Markets, and KBW Research TD as of 4/9/14 Reducing Internalization by Imposing Minimum Levels of Price Improvement: A Trade-At rule (or minimum price improvement) would force more volume back into exchanges by imposing a minimum level at which brokers or ATSs may improve upon the NBBO (for example, a penny better than the NBBO). This was initially proposed in the February 2010 concept release from the SEC and since has also been floated as part of the joint committee with the SEC and CFTC. We think this could significantly impact KCG because we believe that it internalizes a good portion of its wholesale flow in order to maintain profitability and pay for order flow. Other Potential Fallout from Change Proprietary Market Data Revenue Shrinks There is nothing floated right now, but exchanges provide a range of proprietary market data feeds to all customers, including machine readable data, deeper levels of market prices and algorithms that suggest types of buyers and sellers. We estimate proprietary market data makes up 11% to 15% of NDAQ's total net revenue (in 2014), some of which may be more sophisticated level of data only demanded by HFT type customers. Exchanges Eliminate Sophisticated Order Types, Reducing HFT Advantages Exchanges provide many different sophisticated order types, some of which may favor more sophisticated HFT-like customers if all members don t have the resources to program and understand. Maker-Taker Is Eliminated or Scaled Back Many equities and options HFT strategies involve collecting rebates from exchanges, due to the maker-taker pricing structure in these markets. Recently, the WSJ reported that the SEC is considering a pilot program to eliminate maker-taker fees in a select number of stocks for a period of time. If the pilot does not negatively impact retail/institutional investors, spreads, and liquidity, and if maker-taker is subsequently eliminated on a broader scale, we think this could be net-neutral to modestly positive for exchanges' trading fees. However, if there is a significant reduction in HFT due to the eventual elimination of maker-taker on a broader scale in equities, there could be other negative impacts such as a reduction in co-location and market data fees (if this customer group shrinks). The CEO of ICE, who in turn owns the New York Stock Exchange and Arca, has made several public comments challenging the structure of paying for liquidity providers, but neither of its exchanges has yet changed their fees to stop this practice. In the equities markets, the elimination of HFT firms could have minimal impact on trading fees, specifically because most of these HFT firms are collecting rebates from the exchange. However, if regulation was imposed to somehow limit HFT in the futures markets, it could have a larger revenue impact because these customers are fee-paying customers, as the maker-taker Please refer to important disclosures and analyst certification information on pages

14 structure does not exist in futures markets. Impact of HFT on Universal Banks Most Universal Banks will earn revenues from high frequency traders either through trading volumes in company-run dark pools or potential trading volumes as a broker client. However, HFT firms are not typical holders of inventory as market makers so the ancillary security services that banks provide (repos, securities lending, prime brokerage, etc.) are less used by HFT because financing needs are minimal. The Universal Banks will have some potential lending exposure to HFT through lines of credit and loans, but we would view that exposure as small relative to the total lending exposure of these large banks. When we think of the ultimate impact of HFT on markets, the bad players in HFT are usually a broker/dealer themselves and trade directly with the exchanges, in our view. Banks will usually receive very little trading revenues from these firms overall. The bad HFT firms will also be disrupters in market trading (canceled orders, etc.) and that could negatively impact revenues for banks. The largest impact from HFT could be unintended consequences from new market regulation. If regulation resulted in increased bid/ask spread in the equity markets then we believe this would be a large positive for banks. In addition, if regulation forced more HFT into other markets beside equity (FX, commodities, credit) then this could be a negative if it caused spreads to compress and banks see no increase in volumes. The focus on HFT and market structures could cause increased regulation of dark pools as well. In addition, banks do receive rebates from exchanges as a liquidity provider and if a tiered payment structure was disallowed then it could be negative to earnings, all else equal. We view the bottom line impact from HFT revenues to be a small part of bank earnings. However, we believe that HFT has created lost revenues (spread compression, unfilled orders, etc.) and higher expenses (dark pool management, compliance, etc.) so any curtailment of HFT trading could be a positive for banks. In addition, if HFT activity were to increase at broker dark pools due to exchange-driven regulation then this could be a positive for banks. In the end, the most likely impact at banks from potential HFT regulation is increased expense. Expenses may increase if regulators initiate pilot programs to study impacts of HFT and market structures. Pilot programs may not lead to increased revenues in the interim unless regulators actually implement changes market wide. There is a potential that lawsuits related to potential conflicts of interests could emerge as well. We expect any legal issues would be long-tailed but present a new area of risk. We believe that brokers comply with the industry standards for best execution but that still may not prevent lawsuits from occurring. Outside of HFT we view electronic and algorithmic trading as one of the biggest risks to banks due to the potential that errors could occur (i.e., GS s option trading error in 2013). We believe that regulation that reduced the risk of trading errors would be viewed positively by investors. Given that GS and MS are two of the largest brokers in the U.S., we view these companies as having the most potential impact both positive and negative from increased regulation. Regulation focused on retail trading activity would likely impact MS, BAC, JPM, WFC, and GS given the size of each bank s broker network. Since C sold Smith Barney already, we believe the company would be less impacted by retail regulation overall. In addition, both C and WFC have a lower percentage of equity trading revenues compared to total revenues, so both should be less impacted if there were higher HFT-related regulatory costs overall. Please refer to important disclosures and analyst certification information on pages

15 Companies Mentioned in This Report TD Ameritrade Holding Corp. (AMTD: $30.47, Outperform) Bank of America Corp. (BAC: $16.39, Market Perform) Citigroup (C: $48.31, Market Perform) CBOE Holdings, Inc (CBOE: $50.44, Market Perform) CME Group, Inc. (CME: $67.79, Outperform) E*Trade Financial (ETFC: $20.37, Market Perform) Goldman Sachs Group, Inc. (GS: $154.92, Market Perform) Interactive Brokers (IBKR: $21.56, Outperform) Intercontinental Exchange, Inc. (ICE: $193.85, Outperform) Investment Technology Group (ITG: $19.16, Market Perform) JPMorgan Chase & Co (JPM: $54.80, Outperform) KCG Holdings (KCG: $9.72, Market Perform) Morgan Stanley (MS: $29.55, Market Perform) Nasdaq OMX Group (NDAQ: $34.13, Market Perform) The Charles Schwab Corporation (SCHW: $26.11, Market Perform) Wells Fargo & Company (WFC: $48.78, Market Perform) IMPORTANT DISCLOSURES RESEARCH ANALYST CERTIFICATION: We, Joel Jeffrey, Niamh Alexander, Brian Kleinhanzl, Brian Gardner, Michael Needham, CFA and Kyle Voigt, hereby certify that the views expressed in this research report accurately reflect our personal views about the subject companies and their securities. We also certify that we have not been, and will not be receiving direct or indirect compensation in exchange for expressing the specific recommendation in this report. Analysts Compensation: The equity research analysts responsible for the preparation of this report receive compensation based upon various factors, including the quality and accuracy of research, client feedback, competitive factors, and overall firm revenues, which include revenues from, among other business units, Institutional Equities and Investment Banking. COMPANY SPECIFIC DISCLOSURES For disclosures pertaining to recommendations or estimates made on a security mentioned in this report, please see the most recently published company report or visit our global disclosures page on our website at or see the section below titled "Disclosure Information" for further information on how to obtain these disclosures. AFFILIATE DISCLOSURES: This report has been prepared by Keefe, Bruyette & Woods, Inc. ( KBWI ) and/or its affiliate Keefe, Bruyette & Woods Limited ( KBWL ) (collectively KBW ). Both KBWI and KBWL are affiliates of Stifel, Nicolaus & Company, Inc. Keefe, Bruyette & Woods Inc. is regulated by FINRA, NYSE, and the United States Securities and Exchange Commission, and its headquarters is located at 787 7th Avenue, New York, NY Keefe, Bruyette & Woods Limited is registered in England and Wales, no and its registered office is 7th Floor, One Broadgate, London EC2M 2QS. KBWL is authorised and regulated by the Financial Conduct Authority (FCA) in the UK entered on the FCA s register, no and is a member of the London Stock Exchange. Disclosures in the Important Disclosures section referencing KBW include one or all affiliated entities unless otherwise specified. Registration of non-us Analysts: Any non-us Research Analyst employed by a non-us affiliate of KBWI contributing to this report is not registered/qualified as research analyst with FINRA and/or the NYSE and may not be an associated person of KBWI and therefore may not be subject to NASD Rule 2711 or NYSE Rule 472 restrictions on communications with a subject company, public appearances, and trading securities held by a research analyst account. Disclosure Information. For current company specific disclosures please write to one of the KBW entities: Keefe, Bruyette & Woods Research Department at the following address for US Research: 787 7th Avenue, 4th Floor, New York, NY For European Research: The Compliance Officer, Keefe, Bruyette and Woods Limited, 7th Floor, One Broadgate, London EC2m 2QS. Or visit our website at KBW has arrangements in place to manage conflicts of interest including information barriers between the Research Department and certain other business groups. As a result, KBW does not disclose certain client relationships with, or compensation received from, such companies in its research reports. Distribution of Ratings/IB Services KBW *IB Serv./Past 12 Mos. Rating Count Percent Count Percent Outperform [BUY] Market Perform [HOLD] Underperform [SELL] Please refer to important disclosures and analyst certification information on pages

16 Restricted [RES] Suspended [SP] Covered -Not Rated [CNR] * KBW maintains separate research departments; however, the above chart, "Distribution of Ratings/IB Services," reflects combined information related to the distribution of research ratings and the receipt of investment banking fees globally. Explanation of Ratings: KBW Research Department provides three core ratings: Outperform, Market Perform and Underperform, and three ancillary ratings: Suspended, Restricted, and Covered - Not Rated. For purposes of New York Stock Exchange Rule 472 and FINRA Rule 2711, Outperform is classified as a Buy, Market Perform is classified as a Hold, and Underperform is classified as a Sell. Suspended indicates that KBW s investment rating and/or target price have been temporarily suspended due to applicable regulations and/or KBW policies. Restricted indicates that KBW is precluded from providing an investment rating or price target due to the firm's role in connection with a merger or other strategic financial transaction. Covered - Not Rated indicates that KBW is not providing an investment rating and/or price target due to the lack of publicly available information and/or its inability to adequately quantify the publicly available information to sufficiently produce such metrics. North American Stocks are rated based on an absolute rate of return (percentage price change plus dividend yield).outperform represents a total rate of return of 15% or greater. Market Perform represents a total rate of return in a range between -5% and +15%.Underperform represents a total rate of return at or below -5%. European Stocks are rated based on the share price upside to target price relative to the relevant sector index performance on a 12-month horizon. Outperform rated stocks have a greater than 10 percentage point ( pp ) relative performance versus the sector, Market Perform rated stocks between +10pp to -10pp relative performance versus the sector, and Underperform rated stocks a lower than 10pp relative performance versus the sector. The 12-month price target may be determined by the stock s fundamentally driven fair value and/or other factors (e.g., takeover premium or illiquidity discount). KBW Model Portfolio: "Model Portfolio Buy" - Companies placed on this list are expected to generate a total rate of return (percentage price change plus dividend yield) of 10% or more over the next 3 to 6 months. "Model Portfolio Sell" - Companies placed on this list are expected to generate a total rate of return (percentage price change plus dividend yield) at or below -10% over the next 3 to 6 months. The purpose of the Model Portfolio is to inform institutional investors of KBWI s short-term (as described above) outlook for a particular industry sector. The Portfolio is not available for purchase or sale, cannot be duplicated as shown, is hypothetical and is for illustrative purposes only. For a more detailed description of the selection criteria and other specifics related to the construction of the Model Portfolio, please refer to the January 5, 2010 Model Portfolio Primer and/or contact your KBWI representative for more information. The Model Portfolio should be viewed as a short-term outlook of a particular industry sector, not as individual security recommendations. The Model Portfolio uses a three-to-six-month time horizon and should not be considered when making longer term investments. KBWI Research publishes research with a 12-month outlook on each issuer of securities contained in the Model Portfolio. Investors who are interested in a particular security should request KBWI Research s coverage of such securities by contacting your KBWI representative. KBW research contains analyses of fundamentals underlying each issuer. KBWI s long-term recommendations may differ from recommendations made for the Model Portfolio. These differences are the result of different time horizons -- KBWI research has a 12-month outlook and the Model Portfolio has a three-to-six-month outlook. Although the model portfolio is based upon actual performance of actual investments, KBWI did not recommend that investors purchase this combination -- or hypothetical portfolio -- of investments during the time period depicted here. As this hypothetical portfolio was designed with the benefit of hindsight, the choice of investments contained in it reflects a subjective choice by KBWI. Accordingly, this hypothetical portfolio may reflect a choice of investments that performed better than an actual portfolio, which was recommended during the depicted time frame, would have performed during the same time period. Moreover, unlike an actual performance record, these results do not represent actual trading wherein market conditions or other risk factors may have caused the holder of the portfolio to liquidate or retain all or part of the represented holdings. Stifel/KBW Income Opportunity List, Stifel/KBW Analyst Select List, and Stifel/KBW Select Income Opportunity The Crossovers List: These lists include companies that analysts believe have the most attractive risk-adjusted return potential over the next 12 months, or, in the case of the Stifel/KBW Income Opportunity Ideas List, have yield in excess of 2%. In some cases, analysts who cover more than one sub-sector may have more than one name on the list. Other Research Methods: Please be advised that KBW provides to certain customers on request specialized research products or services that focus on covered stocks from a particular perspective. These products or services include, but are not limited to, compilations, reviews and analysis that may use different research methodologies or focus on the prospects for individual stocks as compared to other covered stocks or over differing time horizons or under assumed market events or conditions. OTHER DISCLOSURES Indices: The following indices: U.S.: KBW Bank Index (BKX), KBW Insurance Index (KIX), KBW Capital Markets Index (KSX), KBW Regional Banking Index (KRX), KBW Mortgage Finance Index (MFX), KBW Property & Casualty Index (KPX), KBW Premium Yield Equity REIT Index (KYX); KBW Financial Sector Dividend Yield Index (KDX) and Global: KBW Global ex-u.s. Financial Sector Index (KGX), are the Please refer to important disclosures and analyst certification information on pages

17 property of KBWI. KBWI does not guarantee the accuracy and/or completeness of the Indices, makes no express or implied warranties with respect to the Indices and shall have no liability for any damages, claims, losses or expenses caused by errors in the index calculation. KBWI makes no representation regarding the advisability of investing in options on the Index. Past performance is not necessarily indicative of future results. ETFs: The shares ("Shares") of KBW ETFs are not sponsored, endorsed, sold or promoted by KBWI. KBWI makes no representation or warranty, express or implied, to the owners of the Shares or any member of the public regarding the advisability of investing in securities generally or in the Shares particularly or the ability of its Indices to track general stock market performance. The only relationship of KBWI to Invesco PowerShares Capital Management LLC and ProShares is the licensing of certain trademarks and trade names of KBWI and its Indices which are determined, composed and calculated by KBWI without regard to Invesco PowerShares Capital Management LLC and ProShares, the fund, or the Shares. KBWI has no obligation to take the needs of Invesco PowerShares Capital Management LLC and ProShares or the owners of the shares into consideration in determining, composing, or calculating the Indices. KBWI is not responsible for and has not participated in any determination or calculation made with respect to issuance or redemption of the Shares. KBWI has no obligation or liability in connection with the administration, marketing or trading of the Shares. ETFs trade like stocks, are subject to investment risk, fluctuate in market value and may trade at prices above or below the ETFs net asset value. Investing in a single sector may be subject to more volatility than funds investing in a diverse group of sectors. Brokerage commissions and ETF expenses will reduce returns. In general, ETFs can be expected to move up or down in value with the value of the applicable index. Although ETFs may be bought and sold on the exchange through any brokerage account, ETFs are not individually redeemable from the Fund. Investors may acquire ETFs and tender them for redemption through the Fund in Creation Unit Aggregations only, please see the prospectus for more details. There are risks involved with investing in ETFs, including possible loss of money. Shares are not actively managed and are subject to risks including those regarding short selling and margin maintenance requirements. Past performance is not necessarily indicative of future results. Leverage and Inverse ETFs: ProShares Ultra KBW Regional Banking (KRU) and ProShares Short KBW Regional Banking (KRS) ETFs seek returns that are either 2x, -1x, respectively, the return of an index or other benchmark (target) for a single day, as measured from one NAV calculation to the next. Due to the compounding of daily returns, returns over periods other than one day will likely differ in amount and possibly direction from the target return for the same period. These effects may be more pronounced in funds with larger or inverse multiples and in funds with volatile benchmarks. These funds may not be suitable for all investors and should be used only by knowledgeable investors who understand the potential consequences of seeking daily leveraged investment results. Investing involves risk, including the possible loss of principal. ProShares Short are non-diversified and should lose value when their market indexes rise a result that is opposite from traditional ETFs and they entail certain risks including risk associated with the use of derivatives, imperfect benchmark correlation, leverage and market price variance, all of which can increase volatility and decrease performance. For more on correlation, leverage and other risks, please read the prospectus. An investor should consider the Funds investment objectives, risks, charges and expenses carefully before investing. This and other information can be found in their prospectuses. For this and more complete information about the Funds, call InvescoPowerShares at or visit or call ProShares Client Services at or visit for a prospectus. The prospectus should be read carefully before investing. Shares of the ETFs funds are not guaranteed or insured by the FDIC or by another governmental agency; they are not obligations of the FDIC nor are they deposits or obligations of or guaranteed by KBWI, Invesco PowerShares Capital Management LLC or ProShares. ETFs are distributed by Invesco Distributors, Inc. the distributor of the PowerShares Exchange-Traded Fund Trust II. PowerShares is a registered trademark of Invesco PowerShares Capital Management LLC; ProShares are distributed by SEI Investments Distribution Co. which is not affiliated with ProFunds Group or its affiliates. General Risk Disclosure: Investments in securities or financial instrument involve numerous risks which may include market risk, counterparty default risk, liquidity risk and exchange rate risk. No security or financial instrument is suitable for all investors and some investors may be prohibited in certain states or other jurisdictions from purchasing securities mentioned in this communication. The securities of some issuers may not be subject to the audit and reporting standards, practices and requirements comparable to those companies located in the investor s local jurisdiction. Where net dividends to ADR investors are discussed, these are estimated, using withholding tax rate conventions, and deemed accurate, but recipients should always consult their tax advisor for exact dividend computations. COUNTRY SPECIFIC AND JURISDICTIONAL DISCLOSURES: United States: This report is being distributed in the US by Keefe, Bruyette & Woods Inc. Where the report has been prepared by a non-us affiliate, Keefe, Bruyette & Woods Inc., accepts responsibility for its contents. U.K. and European Economic Area (EEA): This report is issued and approved for distribution in the EEA by Keefe Bruyette & Woods Limited, which is regulated in the United Kingdom by the Financial Conduct Authority (FCA). In jurisdictions where KBW is not already licensed or registered to trade securities, transactions will only be affected in accordance with local securities legislation which will vary from jurisdiction to jurisdiction and may require that a transaction is carried out in accordance with applicable exemptions from registration and licensing requirements. Non US customers wishing to effect a transaction should contact a representative of the KBW entity in their regional jurisdiction except where governing law permits otherwise. US customers wishing to effect a transaction should do so by contacting a representative of Keefe, Bruyette & Woods Inc. ONLY DISTRIBUTE UNDER REGULATORY LICENSE: This communication is only intended for and will only be distributed to persons residing in any jurisdictions where such distribution or availability would not be contrary to local law or regulation. This communication must not be acted upon or relied on by persons in any jurisdiction other than in accordance with local law or regulation and where such person is an investment professional with the requisite sophistication and resources to understand an investment in such securities of the type communicated and assume the risks associated therewith. CONFIDENTIAL INFO: This communication is confidential and is intended solely for the addressee. It is not to be forwarded to any other Please refer to important disclosures and analyst certification information on pages

18 person or copied without the permission of the sender. Please notify the sender in the event you have received this communication in error. NO SOLICITATION OR PERSONAL ADVICE: This communication is provided for information purposes only. It is not a personal recommendation or an offer to sell or a solicitation to buy the securities mentioned. Investors should obtain independent professional advice before making an investment. ASSUMPTIONS, EFFECTIVE DATE AND UPDATES: Certain assumptions may have been made in connection with the analysis presented herein, so changes to assumptions may have a material impact on the conclusions or statements made in this communication. Facts and views presented in this communication have not been reviewed by, and may not reflect information known to, professionals in other business areas of KBW, including investment banking personnel. The information relating to any company herein is derived from publicly available sources and KBW makes no representation as to the accuracy or completeness of such information. Neither KBW nor any of its officers or employees accept any liability whatsoever for any direct, indirect or consequential damages or losses arising from any use of this report or its content. This communication has been prepared as of the date of the report. KBW does not undertake to advise clients of any changes in information, estimates, price targets or ratings, all of which are subject to change without notice. The recipients should assume that KBW will not update any fact, circumstance or opinion contained in this report. COPYRIGHT: This report is produced for the use of KBW customers and may not be reproduced, re-distributed or passed to any other person or published in whole or in part for any purpose without the prior consent of KBW. Please refer to important disclosures and analyst certification information on pages