Thread by @ReformedTrader

1/ Misguided Beliefs of Financial Advisors (Linnainmaa, Melzer, Previtero)

"Advisors trade frequently, chase returns, prefer expensive, actively managed funds, and underdiversify. Their net returns [alphas] of −3%/year are similar to their clients'."

papers.ssrn.com/sol3/papers.cf…

2/ Canadian non-bank advisors and clients (two firms, ≈$20B assets, ≈5% of Canadian Mutual Fund Dealer sector)

Advisors are free to recommend all mutual funds (no captive distribution).

80% of purchases advisors make for their own accounts are also purchased/held by clients)

3/ Clients and advisors display similar trading behaviors with respect to return chasing (using prior one-year returns), very large preference for active management, turnover, idiosyncratic portfolio risk, and fund expense ratios (similar to the Canadian average of 2.41% [!] ).

4/ After fees, clients and advisors have poor performance (-3% alphas after controlling for MKT, SMB, HML, UMD, TERM, DEF).

Advisors get paid to buy funds for their personal accounts. If we remove commissions, they underperform their own clients by 0.6%/year.

Sample: 1999-2013

5/ "Advisor fixed effects capture common variation in fees among clients of the same advisor. These effects substantially increase the model's explanatory power (to 26.1%, indicating that clients who share the same advisor invest in similarly cheap or expensive [fee] funds."

6/ "Including advisor fixed effects significantly boosts the model's explanatory power [for client trading behavior]. The explanatory power does not arise from differences between dealers.

"These estimates suggest that advisors direct their clients to trade in similar ways."

7/ "Even if clients selected advisors who prefer a given strategy, it is unlikely that co-clients would purchase precisely the same funds at the same time without common input from the advisor.

"...strong evidence that advisors direct their clients to trade in similar ways."

8/ "The dependent variable is advisor a's estimated fixed effect for trading behavior i from the analysis reported in Table 5.

"An advisor's personal investment behavior correlates closely with that of his clients."

9/ "An advisor's clients often buy the same new fund as the advisor within a few months of the advisor's own purchase.

"There is little overlap in purchases with respect to clients of other advisors" (whether they are from the same dealer or a different dealer).

10/ Clients who switch advisors tend to buy similar funds as whichever advisor they happen to be working with at the time.

"This pattern is consistent with a causal connection: advisors' preferred investments appear in clients' portfolios specifically while they work together."

11/ "Funds purchased only by advisors [1/5 of total advisor purchases] have higher prior returns, more idiosyncratic risk, and higher fees.

"The differences between client-only and joint purchases [by client and advisor], by contrast, are small."

12/ Advisors with the worst performance in their personal portfolios also have clients with the worst performance, though EVERY group underperforms the market.

The difference between best and worst is highly significant (t-stat>5).

The "best" advisors have the fewest clients.

13/ "Clients do not predominantly gravitate toward advisors who invest in cheaper, better performing and more diversified portfolios. Rather, the advisors who buy high-cost and underdiversified funds advise, if anything, more clients than their peers."

14/ "Advisors do not substantially alter their investment behavior after they quit the industry.

"Expense ratios decrease, but this change reflects an increased allocation to fixed income: the within-asset class fee remains nearly unchanged at the 46th percentile."

15/ "In contrast to the strategic trading conjecture [buying poorly performing funds to convince clients to do the same], the correlation in advisor-client trading is 2x-3x larger for advisors with the largest relative portfolios [portfolios that are not 'play money']."

16/ "Advisors could significantly improve their performance by holding the same portfolios as their clients. Poorly performing funds do not appear just among investments held jointly with clients but are actually more prevalent among investments made by the advisors alone."

17/ "The typical mutual fund underperforms because it fails to cover its fees.

"Investments recommended by financial advisors, by contrast, are poor investments even before commissions.

"Clients would underperform the market even if advisors provided services free of charge."

18/ "Advisors' personal trading behavior is stable throughout their careers and not just when they stop advising clients.

"Adivisors typically have positive net inflows, so their portfolio values increase; turnover therefore decreases over time as the denominator increases."

19/ Investors who later become advisors don't substantially change their trading behavior (in personal accounts) as a result.

The stability of their trading behavior pre/during/post-career suggests that their beliefs are not affected by their level of job training/experience.

20/ "Regulations that reduce conflicts of interest (imposing fiduciary duty, banning commissions) do not address misguided beliefs.

"When advisors recommend strategies that underperform, they act as an agent exactly as they do as a principal [in their personal accounts]."