Right around Labor Day, the sleepiest time in mortgage markets, Institutional Investor published it's All American Fixed Income Research Team results
Notably missing in the survey's mortgage-related "Structured Securities" categories was Laurie Goodman, who has been a senior managing director at Amherst Securities since November 2008. Previously she was at UBS
, where she and her team of MBS analysts regularly dominated the survey.
Yes, I'm going to complain, and yes I have a bias: It's well known that I've worked for Goodman off-and-on for about nine years over the course of my research career, and she's been a valued colleague, whether a boss or a competitor. (Goodman declined to be interviewed for this commentary.) But other fine mortgage analysts whose work I avidly read and admire, were also neglected in the survey.
Goodman and the others are absent from the published results because Institutional Investor's fixed-income research survey methodology is seriously flawed, particularly so with regard to securitized product. Part of the problem is that Institutional Investor bases its vote count on the Securities and Exchange Commission's Reg AC, without understanding how poorly Reg AC's definition of "research" fits the activities and products of mortgage and asset-backed analysts.
Biased or not, in my research career I always suspected the fixed-income survey might be more hoopla than an actual benefit to analysts or investors. More to the point, I never thought Reg AC made a useful contribution to the supervision of fixed-income research, and the efforts of some firms to conform their policies for fixed-income analysts to those mandated for equity analysts have always struck me as wrongheaded constraints on commerce. Finally, I have come to wonder if the surveys themselves are not uncomfortably reflective of an increasingly oligarchic bond business. How Goodman fared in the poll just gives me a great excuse to lay out my arguments in public.
How fickle is fortune
At UBS, Goodman was co-head of fixed-income research globally and head of securitized product research in the U.S. There, she and analysts working for her dominated survey results with team finishes in virtually all mortgage-related categories for over a decade. In particular, she was named top MBS strategist every year from1998 through 2008. That changed with the abrupt reorganization of the securities industry by the financial crisis. Investors are surveyed in the spring, results published in the fall. By the time the UBS team was celebrating its best performance ever, in 2008, with nine first or second team spots, UBS had begun to dismantle the nonagency mortgage, asset-backed securities and collaterized debt obligation businesses that Goodman's research team supported. Goodman moved to independent broker-dealer Amherst late in 2008. Since it takes months to bring online the human and technological resources needed in mortgage research and to knit a new research group into a burgeoning sales and trading effort, no one was that surprised when her only mention in the Institutional Investor analyst survey a few months later, was runner-up in MBS strategy. And in the grand scheme of things, it wasn't that important — 2009 was the year she was inducted into the prestigious Fixed Income Analysts Society's Hall of Fame.
But the 2010 Institutional Investor survey results — again, runner-up in a single category, MBS Strategy — are baffling. Maybe runner-up in a broad strategy category makes sense — Goodman's focus has narrowed to reflect the market footprint of her new firm — a "regional" broker-dealer that doesn't do soup-to-nuts bond and derivatives trading and doesn't need a research department that can be all things to all people.
What makes no sense is her exclusion from the structured securities product areas — MBS/Agency Structured Products, MBS/Nonagency-Structured Products, ABS/Real Estate or MBS/Prepayments — where she has drilled deeply, and for many readers of her publications, struck gold. In the last year, Goodman's work has had broader exposure and received bigger kudos from a vastly larger audience, thanks to the unfinished crisis, than any year since she began publishing fixed-income research in the 1980s. She's been called to testify before the House Financial Services Committee, she's been on Bloomberg TV
and PBS' Nightly Business Report. Her speaking engagements include the likes of the Milken Institute, the National Economics Club, the FDIC and the Professional Risk Managers' International Association. At its invitation-only housing finance conference, AmeriCatalyst devoted an entire session to Goodman, who detailed her analysis of shadow housing inventory and loan modification programs.
Methodology and technicalities
Other excellent analysts whose work I admire were snubbed as well. The problem isn't with the analysts, it's with the survey methodology. It is very possible the analysts got the votes but they weren't counted. I can't speak for the other good analysts, but I've been told that Goodman's name did appear on the ballot for those mortgage-related products she supports at Amherst.
Looking back, I have similar doubts about the 2009 results as well. It appears — eyeballing the results — that one of my favorite CMBS analysts, Lisa Pendergast, at RBS at the time the votes were collected, was denied credit for her runner-up showing. Looks like it was given instead to a general strategist at the same firm (I found the results via Google.
Those votes for analysts like Pendergast at RBS or Goodman at Amherst were tossed out on a technicality. Here's how II explains it: "Analysts must be certified pursuant to Regulation AC to be recognized as winners or team leaders in research sectors (Investment Grade, High Yield, Structured Securities); certification is not required for analysts and team leaders to be recognized in Strategy & Economics sectors.
" (This methodology was disclosed after the fact, in discussion of the results. It was not, I've been told, disclosed on the ballot.)
The definition of "research"
Institutional Investor has chosen to use the SEC's definition of "research" (as promulgated in "Reg" AC) — which, like virtually all registration and regulation of security analysts at broker-dealers, as well as of their publications and public appearances, is based on the equity research model.
Equity research practices and outputs share many commonalities with corporate bond research, especially in the high-yield market, but virtually none with what is colloquially meant by research in securitized products.
I'll elaborate, but first, a technical quibble of my own. I think Institutional Investor has misread Reg AC. Reg AC does not require certification of analysts. It requires authors of what Reg AC defines as "research," who are associated with broker-dealers to certify that their research expresses their own personal opinion. Also, they must certify that, either their compensation was not based on the specific analysis and recommendations or, if it was, say so and state the amount and source of the compensation.
Again, Reg AC does not say anything about designating authors of "research" as Reg AC analysts. It assumes that the potential for conflicts of interest is inherent. The onus is on the analyst to assert that the so-called research was not influenced by conflicts of interest. Other sources of influence — promotions, nice offices, outings, verbal pressure, whatever — are not at issue.
Research narrowly defined
Reg AC defines what kinds of publications or public appearance are "research" and hence must be certified. For instance, "Research report means a written communication (including an electronic communication) that includes an analysis of a security or an issuer and provides information reasonably sufficient upon which to base an investment decision."
In the adopting release, the SEC supplied broader guidance. Certain communications would not be research reports so long as they do not include analysis of, or recommend or rate, individual debt or equity securities or issuers, such as reports discussing broad-based indices such as the S&P 500; reports commenting on economic, political or market conditions; reports commenting on or analyzing particular types of debt securities or characteristics of debt securities; technical analysis concerning supply and demand for a sector, index or industry based on trading volume and price; and reports that recommend increasing or decreasing holding in particular industries, sectors or types of securities.
That is, more or less what is commonly called strategy. Anyone who has read or written asset-backed or mortgage-backed research knows that the weekly and special reports are not written to encourage investment in a specific security or issuer. They focus on types of securities, security structures or sectors, relative value, market conditions — all the topics the SEC does not consider to be research. However, particularly in the case of mortgage securities, it is rarely convenient — or even possible — to discuss a type or sector without using an example security to illustrate the effect of different interest rate/prepayment scenarios on performance. Every analyst I know — and all the traders who ever assisted me in this task — takes great care to select generic, representative
examples, but all that care doesn't change the fact that, according to Reg AC, including a realistic example could qualify the discussion as research.
If it smells like research
So what if the report crosses the line? Shouldn't an MBS analyst still be able to make the necessary certifications? It comes down to the reporting lines and who makes the compensation decisions. If an analyst, or the analyst's boss, reports to sales or trading, then any assertions that the analyst isn't compensated for the specific recommendations may be true, but easily challenged. Likewise, some of the best, most interesting and satisfying ideas I ever had for reports came from traders. A few came from sales folk who were great conduits of their customers' interests and questions. Did I believe in my heart every word of the resultant report? Yes! Could someone challenge and say, that's the same thing your head trader said when we were playing golf last week? Yes!
Readers should conclude that analysts who support sales and trading of securitized assets, no matter how independent, are not in a position to make Reg AC certifications if they also report to sales or trading. Big firms can solve this problem. They have enough boxes in the org chart to hang research one or two lines away from the revenue stream. Nonetheless, make no mistake, sales and trading still control the size of the pay pot, have lots of veto power over anything to do with their businesses and will make sure that analysts whose efforts they deem to be valuable are compensated appropriately.
Smaller firms do not have the "management economics" to create this illusion that fixed-income analysts aren't compensated by sales and trading. But even firms that do have boxes that could facilitate Reg AC certifications may still not consider it worth their while to impede the flow of information between analysts and traders by creating such illusions.
Finally, and more importantly, analysts are not dumb. They know they get paid from trading profits, regardless of how they are laundered through the organizational chart, and they know that what they write should support trading flows so they can get paid.
Alternative — Don't call it research
Many firms decided to scrap the notion of publishing Reg AC certified research, rather than bulk up the organization with extra management, compliance officers, rule writers and what not, in order to make nonsensical assertions regarding the independence of views. Instead, they don't call it research. Typically, the folks who maintain the databases, prepayment and pricing models and spread, price and prepayment reports are trading desk employees. People who generate trade ideas, commentary and reports are desk analysts. These published reports are identical in tone, content and level of detail to the ones that their counterparts put Reg AC testimonies on. The reports still carry disclosures to remind customers that the firm has its own monetary interests that may not necessarily be aligned with those of its customers.
The difference is that the reports are labeled "material" or "information." The disclosure states they are prepared by sales and/or trading personnel. It may even be explicitly stated that the reports are not research produced by a research department. This is the kind of disclaimer made, for example, by Amherst, First Tennessee and RBS.
But the material produced by these analysts satisfies all the same purposes and is as relevant to the current trading environment as research. Investors who rely on the material take it just as seriously as "research."
But who reads such disclosures? Until I started trying to puzzle out what happened in the 2010 fixed- income survey, I never looked at them. So I doubt that folks who like Amherst, FTN, Jeffries or RBS' structured securities research, for example, ever knew the SEC was drawing a line between that research and "real" Reg AC certified research. And it wasn't mentioned on the Institutional Investor ballot (I've been told it wasn't), they wouldn't know the distinction determined whether their votes counted or not.
Reg AC just a piece of the puzzle
Institutional Investor's editors might not have made Reg AC a bright-line test for their fixed-income survey, if they had a sense of how complex and patch-worked the rules are for making transparent the conflicts of interest that naturally exist between security analysts and the broker-dealer businesses. Or that they are particularly patch-worked and makeshift when applied to fixed income because they were designed for equity markets, where analysts do shill for the shares of individual companies.
In fact, they were designed for equity markets because that's where analysts did publish research, for which they were grandly compensated, that promoted securities they privately considered to be worth less, if not worthless (as, for instance, internal e-mail traffic attested). A decade ago, these and other abuses of the research function resulted in high-profile investigations by the New York Attorney General and the SEC and splashy congressional hearings. The scandals resulted in rule amendments in 2002 by the New York Stock Exchange and the National Association of Securities Dealers designed to address equity research analyst conflicts of interest. The same year, Sarbanes-Oxley was enacted, including a section amending the Securities Exchange Act of 1934 to address the conflicts of interest that may arise when securities analysts recommend stock in equity research reports and public appearances. In 2003, the SEC adopted Reg AC and the self-regulated organizations implemented additional restrictions on equity analysts mandated by SOX changes to the Exchange Act.
Also, in 2003 enforcement actions resulting from the investigations were finalized.
In addition to extracting almost $1.4 billion of penalties, disgorged profits and other payments from 10 very big firms, the settlements required the firms to make significant structural changes to ensure that "stock recommendations are not tainted by efforts to obtain investment banking fees." The changes included requiring physical separation of research and investment banking departments (lots of floors were "restacked"), and requiring senior management to set research department budgets without input from investment banking and without regard to specific investment banking revenue. Furthermore, investment banking was to have no role in decisions regarding analyst compensation or evaluating their job performance or in decisions regarding company-coverage decisions.
The same structural changes were generally incorporated in the SROs' rule changes adopted over the same period.
It's all about equity
Please note, of all the rules coming out of those scandals and meant to assure analysts' independence, that ONLY Reg AC applies to fixed-income, as well as equity analysts.
Also note that the structural changes resulting from the equity research scandals deal with the relationship between equity analysts and investment banking. Trading and sales functions are not explicitly mentioned, though they are implicated because they do directly and indirectly compensate analysts. This is the legal patchwork that has inspired the legal departments of some broker-dealers to build out separate management structures for fixed-income research, others to relabel research, and others to stop publishing it entirely.
Too big to fail wins again
If you've read this far you may be scratching your head, why should you care? Institutional Investor charges $400 a year for a subscription, assuring a small audience of higher-ups in investment management and securities firms. Its rankings are a main source of content in the magazine and a leading motivation for subscribers. It also collects additional data that is not a part of the rankings that it can sell.
In other words, it's commercialized. Conveniently, it gives broker-dealers an independent measure of performance on which to base analyst compensation (since in fixed income it's hard to directly relate research expense to trading profits). It tells headhunters where to look. It influences marketing and market positioning.
The poll is yet another force (hopefully weak) for consolidation and concentration, for the creation of too-big-to-fail, systemically risky institutions. First of all, the methodology favors size. Each vote is weighted by the investment manager's fixed-income assets under management. One might imagine they drowned out the votes of investment boutiques. Not that dollars under management don't in some way reflect a firm's competence, but in many cases they are more of an indication of M&A power, the ongoing consolidation of every segment of the financial services industries.
Taken together, the results reflect the ongoing concentration of banks and security dealers. Although 18 firms received enough votes in 2010 to at least qualify as runners-up in more than 60 categories, just vie firms — J.P. Morgan, Barclay's Capital, BofA/Merrill, Goldman and Wells, everyone of them a too-big-to-fail financial enterprise — took 80% of the first, second or third team and runner-up positions.
This accords with other industry statistics. The number of primary government security dealers
has collapsed from 38 in 1992, 29 in April 2000, 23 in August 2006 to just 18 now.
Likewise, the number of brokerage firms overseen by the Financial Industry Regulatory Authority, formerly the National Association of Securities Dealers, has dropped from 5,111 in 2005 to 4,657 as of August.
Change the procedure
I am not calling for the end of II surveys. They have entertainment value just like the lists of best and worst dressed ladies at the Oscars, or top 100 rock and roll singles of all time. But if the folks at Institutional Investor want to deserve to be taken seriously, they need to align the rules with investors' expectations, not regulatory mumbo-jumbo.
Maybe they thought they took a step in that direction with the 2010 poll. They killed CMBS and CDOs (Collaterized Debt Obligation) as product sectors. And added CMBS as a strategy sector. But if they were thinking CMBS and CDO are dead, they were premature. JP Morgan hit the headlines in the last week or so with a big CMBS deal. And I continue to read about CLOs (Collateralized Loan Obligations), a CDO sub-sector.
The better move would be to ignore the Reg AC distinction, and include all analysts who publish regular structured securities research. Or don't just open up CMBS, instead call them all strategy sectors.
Tighten the categories so they reflect what people are looking at: Prepayments and credit in agency MBS, jumbo and Alt-A MBS and subprime MBS. That could almost be six different categories. Get rid of that dopey ABS/Real Estate label. A mortgage is a mortgage is a mortgage. Calling subprime mortgages ABS was a marketing ploy after the CMO market crashed in 1993 and the concept of tranched MBS was tainted. It's old and artificial and makes the publication's editors look out-of-touch. Find a way to recognize that the very finest research on the housing market is coming from the mortgage analysts' sophisticated credit work. Likewise, a lot of the very best stuff I have read about the regulatory environment has also come from fixed-income analysts — mortgage securities analysts and agency security analysts in particular. Why not establish a category in fixed income for Demystification of Regulatory and Accounting Rules