Abacus Analytics is pleased to announce the completion of development work on a new strategy backtesting system. Berry Cox, founder and principal of Abacus, is the architect of this new capability.

"We view strategy backtesting as a logical extension of our work in research evaluation," said Cox. "Our work in analyst evaluation confirms that many research organizations have an impressive ability to separate outperforming stocks from underperformers. Stock-picking ability, however, does not translate directly into value-added portfolio management."

"One obvious reason for this is that most research departments' Buy lists, even when they add value, are far too long to become real portfolios. In addition, portfolio managers need to pay attention to diversification and risk levels and to turnover constraints in a way that security analysts do not."

"Our backtesting work starts by evaluating the statistical properties of a manager's underlying investment recommendations to confirm that they are, in fact, alpha generators. Based on the properties of a client's signals, we determine which strategy parameters are more likely to maximize out-of-sample, real-time performance. The Abacus system evaluates diversification levels, turnover and risk constraints, rebalancing frequency, sell disciplines, stop-loss rules, and other portfolio construction techniques. Our turnaround time for these analyses is measured in days rather than months," said Cox. "This makes us ideal partners both for investment managers who want to refine their current strategies, and for research providers who want to translate their research recommendations into actual managed portfolios."

Abacus Analytics is pleased to announce a major new functionality in its Analyst Evaluation Service. Berry Cox, founder and principal of Abacus, discussed the firm's new Alpha Accrual reporting.

"Abacus' business has always involved helping research managers evaluate the predictive ability of their analysts. Our clients have had many questions about how this ability manifests itself over time. For example, we've noticed, in a number of cases, that recommendations with a relatively short duration (say less than a quarter) tend to outperform recommendations with a relatively longer duration (say more than a year)."

"To understand this effect more completely, we've created a set of event studies that we call Alpha Accrual reports," Cox explained. "These reports measure how quickly an analyst's insights get incorporated into market prices. Imagine a research process in which the average Buy recommendation outperforms the market by 5%. Alpha Accrual lets us understand how quickly that 5% appears. For some research processes, most of the gain comes very quickly, perhaps in the first 90 days. For others, it may take much longer for the benefit of the analyst's call to be apparent."

"If research managers understand where on this continuum their analysts fall, the potential to improve performance is obvious."

Commentators often refer to a weak stock market environment, like today's, as a "stock picker's market." We're not particularly comfortable with that designation, if for no other reason than its implication that in some environments, stock picking is irrelevant. We've never encountered either an active portfolio manager or a provider of equity research who accepts that view. The accompanying note provides a simple test of the notion of a stock picker's market by comparing measures of analysts' stock-picking performance in both strong and weak equity environments over the past 24 months.

We believe that stock selection can be a source of analyst value added regardless of the market's overall trend. In this sense, there are no stock picker's markets. There are, however, stock picker's metrics - i.e., metrics that can help us discern an analyst's ability to separate winners from losers, without being confused by the market's directional trend. Understanding and using these metrics is the key to identifying analysts who can add value on a sustainable basis.

View Note

PDF version is available here.

Earlier this week, The Wall Street Journal published its annual "Best on the Street" ranking of sell-side analysts. Like most league tables, the Journal relies on a fairly simple methodology that can be easily run for many analysts at many different firms. Such rankings can be informative, up to a point. We suggest in the accompanying note that it's important for research managers to consider a variety of metrics, because they typically have a variety of questions about the performance of their analysts. Otherwise the question of who's number one can depend too much on whose number one uses.

View Note

PDF version is available here.

Yesterday's Wall Street Journal carried its annual "Best on the Street" ranking of securities analysts, which purports to identify 2006's best stock pickers. As the accompanying note points out, there's more to the Journal's rankings than stock picking. Moreover, and more importantly, we believe that it's important for research managers to distinguish between their analysts' abilities at coverage timing and stock selection. Although both can be important sources of research value added, their sustainability - and their value to the buy side client base - are likely to differ.

View Note

PDF version is available here.

Providers of sell-side analyst evaluation services sometimes try to summarize ratings performance with a single number. We've long been critical of such one-factor approaches, largely because we believe that no single metric can adequately capture every dimension of an analyst's job. In fact, the problem with the single-factor approach is not just that it overlooks some subtle complexities - there are circumstances in which such an evaluation technique can actually be misleading.

The attached note posits an example in which a single-factor analyst evaluation scheme gives us the wrong decision when trying to assess which of two competing analysts has done a better job. This simple example makes an important point: there are several dimensions of analyst skill, and they are not all equally important to the buy-side customer base.

View Note

PDF version is available here.

Yesterday it was announced that Fidelity Investments had agreed with Deutsche Bank to unbundle the pricing of research and execution services. This news follows a similar deal between Fidelity and Lehman Brothers earlier this year. A number of commentators have suggested that these two agreements presage the inevitable separation of research and execution, with ominous implications for the future viability of Wall Street and independent research.

We do not share this view, for two reasons. First, we view Fidelity's initiative as an attempt to gain competitive advantage relative to other buy-side firms. It's too soon to say whether that advantage will be real or illusory, and far too soon to be certain that other buy-side firms will imitate Fidelity's move to unbundling. Second, regardless of how widespread unbundling becomes, the economics of sell-side research began to change several years ago with the loss of the corporate finance subsidy. We believe that research departments will survive and prosper to the extent that they can systematically demonstrate and market their ability to add value to client portfolios.

View Note

PDF version is available here

Abacus Analytics is pleased to announce the introduction of a new analytic platform for its Analyst Evaluation Service. Berry Cox, founder and principal of Abacus, commented recently on some of the features of the new system:

"Our analytic work has always been organized around four dimensions of ratings performance: Profitability, Breadth, Reliability, and Timing. Profitability measures the degree to which a user could have made money by owning a research provider's Buy-rated stocks and shorting his Sells, in both absolute and risk-adjusted terms. Breadth and Reliability are both measures of consistency . Breadth measures consistency across stocks, and Reliability measures consistency over time. Consistency is perhaps more important than profitability per se. That's because buy-side customers consume research one recommendation at a time. The greater a research provider's consistency, the more likely it is that managers will add value (alpha) with their use of the research."

"Finally, Timing is as aspect of ratings evaluation that's often overlooked, because it's not designed to measure a research provider's stock selection ability. The idea of our market timing metrics is to identify whether the distribution of an provider's ratings tells us anything about the future direction of the market or the analyst's sector."

"In the new analytic platform, our ratings portfolios reports devote a distinct section to each of these four dimensions of ratings performance. That makes it easier for our clients to home in on precisely the measures they're interested in. The additional size of the report lets us provide a much more comprehensive picture of each aspect of performance."

"One of the most important aspects of the new system is its incorporation of investment style data into ratings evaluation. Abacus is uniquely positioned to do this given our work in portfolio analysis and alpha generation for buy side clients. Our objective is to determine if particular stylistic elements inform a firm's, or an analyst's, ratings decisions. For example, does an analyst tend to upgrade stocks after they've risen (which would indicate a momentum bias) or after they've fallen (which would indicate a value bias)? Understanding these tendencies gives us insights into analyst behavior . and more importantly, may help us not only to report on, but potentially to help improve, our clients' performance."

In late October, Fidelity Investments and Lehman Brothers announced what is described as a "pilot program" to separate payment for research and execution services. As and if such unbundled structures become more common, the competitive environment for both Street and independent research providers is likely to become tougher, as the pressure on research to demonstrate its value to the institutional buy side increases. In this sense, unbundling represents a continuation of the trend that began with the separation of research and investment banking in the global research analyst settlement.

If the bad news is that Street research must stand on its own, the good news is that much research can stand on its own. The buy side is willing to pay for research that provides value, i.e., that helps portfolio managers achieve their performance objectives. Research salespeople will increasingly be required to demonstrate that value. Understanding their analysts' predictive abilities will become a necessary component of the research sales job.

This note suggests some ways in which systematic analyst evaluation can make the research sales process more effective.

View Note

PDF version is available here.

Recently the British Financial Services Authority issued a policy statement and amended regulations on the use of commission revenue to buy research services. The direct impact of the new rules, which limit the use of commissions and require periodic client disclosure, is on the U.K. investment management industry. We suspect that their indirect impact, on brokers and other providers of investment research, will be even more important. In order to qualify for purchase with commissions, investment research must be "capable of adding value" to an investment manager's decision process. The obvious question, for providers of investment research in the U.K. market, is how to demonstrate that their research is "capable of adding value". If regulators elsewhere follow the FSA's lead, in whole or in part, the question becomes even more critical.

View Note

PDF version is available here.

This morning's Wall Street Journal reported on the results of its 13th annual ranking of Wall Street security analysts. We prepared this note in response to a client inquiry, and thought that you might find it useful as well.

View Note

PDF version is available here

Abacus Analytics is pleased to announce its relocation to

372 Danbury Road
Suite 181
Wilton, CT 06897.

“This move is a natural step in the growth of our business,“ said Abacus founder Berry Cox. “We've outgrown our existing office space, and are pleased to be moving to new facilities.“

372 Danbury Road, also known as Wilton Plaza, is a class A office building located approximately 8 miles north of Abacus' former office in South Norwalk, Connecticut.

Abacus Managing Director Craig Lazzara has co-authored a chapter in The Investment Think Tank: Theory, Strategy, and Practice for Advisers, published today by Bloomberg Press. The chapter, co-authored with Gary L. Gastineau, is called "Reinventing the Investment Fund."

"Gary Gastineau and I worked together at ETF Advisors before I joined Abacus," Craig explained. "This chapter is a result of that collaboration, and suggests ways in which conventional mutual funds and exchange-traded funds might evolve to serve investors more efficiently."

The Investment Think Tank was edited by Harold Evensky and Deena Katz, two of the nation's most-respected financial advisers, and includes contributions from many prominent advisers and academic experts.

Abacus Analytics is pleased to announce that Craig J. Lazzara has joined the firm as a Managing Director. He will be responsible for client relations and marketing of Abacus suite of research and consulting services.

Berry Cox, Abacus founder and principal, said that he and Craig had originally been colleagues at Salomon Brothers, where they worked together in the mid-1990s. “Craigs joining the firm strengthens our team significantly,“ Berry said. “We are both pleased to be working together again.“

Prior to joining Abacus, Craig directed marketing and client service for ETF Advisors, a sponsor of exchange-traded funds, and for Salomon Smith Barneys Global Equity Index Group. Earlier in his career, Craig was chief investment officer of Centurion Capital Management and Vantage Global Advisors, where he managed quantitatively-disciplined equity, tactical asset allocation, and currency management strategies. He also served as a managing director of TSA Capital Management, with responsibilities for both applied research and client relations, and as a vice president and portfolio manager for Mellon Bank and T. Rowe Price Associates.

Abacus Analytics has launched its new website, www.abacus-analytics.com. The firm intends that site will become the preferred delivery method for most client reports and updates.

Berry Cox, Abacus founder and principal, commented that a typical months reports for a client of the Abacus Analyst Evaluation System will require reams of paper, and that delivery of electronic data, while less physically cumbersome, still requires extensive and time-consuming internal processing. “Web delivery is more efficient, and clients can see their updated reports at the same time that we have them here,“ he said.