Last week we posted Have 1.3% or 7.3% of Stock Brokers Engaged in Misconduct? explaining that the competing estimates of broker misconduct differ because of differences in their definition of misconduct and the sample of brokers studied.
Firms with 400 to 999 Brokers Are Much Worse Than Larger Firms.
In last week's post, we listed the 100 highest risk brokerage firms with 400 or more registered brokers sorted by the percentage of their brokers associated with Investor Harm Events as of December 31, 2015 as defined by the FINRA study, "Do Investors Have Valuable Information about Brokers?". The 10 highest risk firms with more than 400 brokers, listed below in Figure 1, are the same 10 firms whether using the FINRA study definition or Egan, Matvos and Seru's broader financial misconduct measure from "The Market for Financial Adviser Misconduct".
Figure 1. 10 Highest Risk Firms with More Than 400 Brokers, December 31, 2015
Nine of these 10 firms - Aegis Capital, Summit Brokerage Services, National Securities, Centaurus Financial, Independent Financial Group, Kovack Securities, Wedbush Securities, Investors Capital Corp. and Wunderlich Securities - have between 400 and 999 brokers and so are not included in Egan Matvos and Seru's Table 6 reproduced here as Figure 2. That is, 9 of the 10 highest risk firms with more than 400 brokers have fewer than 1,000 brokers. And the worst of these firms with more than 400 brokers are twice as bad as the worst firms in Egan Matvos and Seru's Table 6 which has brought so much attention Oppenheimer, First Allied and Wells Fargo Advisors.
Figure 2. Egan Matvos and Seru's, page 37, Table 6
The Smaller Firms Are Even Worse!
The brokerage firms with fewer than 400 brokers are even worse than the brokers with between 400 and 999 brokers. Figure 3 below relaxes the threshold to include firms with 100 or more brokers. There are 32 higher risk brokerage firms with 100 or more brokers before we get to Oppenheimer sorted by brokers with Investor Harm Events. 26 of these 32 higher risk firms have fewer than 400 brokers each. On average, 17.9% of the brokers at these 26 firms with between 100 and 399 brokers have had at least one Investor Harm Event compared to 17.0% of the brokers at the six firms ranked higher risk than Oppenheimer with more than 400 brokers.
Figure 3. 42 Highest Risk Firms with More Than 100 Brokers, December 31, 2015
It's Ethics, Not Product Failure
UBS Financial Services of Puerto Rico has the highest percentage of brokers registered as of December 31, 2015 associated with an Investor Harm Event in Figure 3. This is perhaps not surprising given the large number of customer complaints filed over UBS Financial Services of Puerto Rico's closed end bond funds. (You can read our February 12, 2015 summary blog post, "UBS Puerto Rico's Bond Fund Debacle: What We Know So Far for more information.)
Some commentators have incorrectly asserted that assessing brokers based on customer complaints is unfair because it tars good brokers with the results of a failed product promoted by their employer. While this might occasionally happen, it is surely the exception not the rule.
First, the rankings in Egan Matvos and Seru's Table 6 covering firms with more than 1,000 brokers and our tables covering brokers with more than 400 or more than 100 brokers identify firms not brokers. To the extent a large number of brokers received customer complaints because of a faulty product promoted by their firm, the identification of these firms as high risk is a service to investors.
Also, even at firms which have been identified with a large product failure like Morgan Keegan with the RMK Funds, Citigroup with MAT/ASTA and UBS with the Lehman Brothers structured products the vast majority of brokers were not subject to customer complaints. If a broker has a lot of customer complaints related to a product failure the broker likely bears some significant responsibility.
Finally, brokers with a customer complaint in one year are far more likely to have customer complaints in the next year, the second year, the third year and so on. If the extent of customer complaints was related in any meaningful way to product failures, we wouldn't see the persistence we observe in the data.
