Labor Department Wants to Tweak Your Retirement Plan
Yesterday, the Department of Labor released its proposed rules "to protect consumers from conflicts of interest in retirement advice" by requiring brokers to act in their customers' best interests when dealing with individual retirement accounts. The concern is that a lot of retirement advice is provided by brokers who receive payments from mutual fund companies for recommending those companies' funds, and so recommend worse products to investors than they would if they were solely looking out for the investors. One classic American way to solve that problem would be by requiring the advisers to tell the investors about the conflicts, but the Labor Department worries about that approach:
A large body of research has found that the effects of disclosures by themselves are limited and, in some cases, can lead to harms and weaker consumer protections. Indeed, many financial advisers already provide disclosures and the evidence suggests that they are not highly effective -- consumers rarely understand how their advisers are regulated or paid, and can seldom effectively guard against the impact of conflicts, even when the conflicts are disclosed. 1
A more direct way to solve the problem would be to prohibit brokers from taking the payments from mutual fund companies that create the conflicts of interest. But the Labor Department doesn't like that approach either:
The Department believes the more-tailored approach reflected in its proposal can significantly reduce the harms created by conflicts of interest, while continuing to allow for common forms of compensation.
It's odd, right? The "common forms of compensation" -- commissions and revenue sharing from the companies providing the products -- are what the Labor Department is complaining about. If the conflicts of interest weren't caused by "common forms of compensation," you wouldn't need a new rule. But the Labor Department will continue to allow the common forms of compensation that create conflicts. Advisers will just need to -- guess what! -- disclose them.
I mean, there's more to it than that: The adviser has to enter into a "best interest contract" with the client promising to act in the client's best interest, adopt policies "designed to mitigate conflicts of interest," and "Clearly and prominently discloses any conflicts of interest, like hidden fees often buried in the fine print or backdoor payments, that might prevent the adviser from providing advice in the client's best interest," including by referring "the customer to a webpage disclosing the compensation arrangements entered into by the adviser." But you'd expect that if the Labor Department was worried about common practices that create conflicts of interest, it would ban those practices. Instead it will just dress them up more.
That said, there are two specific things that the Labor Department seems to be getting at: Mutual fund loads, and active management. Mutual fund loads are the somewhat old-timey way of marketing mutual funds in which you pay the fund manager a big upfront fee when you buy a mutual fund, and the fund manager pays most if it straight back to the broker who sold you the fund. So the higher the fee, (1) the worse off you are (you paid it) but (2) the better off the broker is (he gets it). This is a straightforward enough conflict of interest, albeit one that exists in all sales relationships, but it's worse than it sounds. Here is a boring-looking table from the Regulatory Impact Analysis that feels like the centerpiece of the Labor Department's rulemaking:
The average up-front cost of a fund with a front-end load is 1.8 percent of the amount invested, the large majority of which goes not to the mutual fund company to pay for its research or whatever, but to the broker for having recommended the fund to you. But there's also the "effect of loads on returns": For every extra 1 percent that you indirectly pay your broker for recommending a fund to you, the fund on average performs 0.45 percent worse than a cheaper fund.
The reasoning is pretty intuitive: Funds can sell themselves by marketing performance, or by relying on broker kickbacks, and the more they invest in broker kickbacks the less they need to invest in performance. The Labor Department cites a paper supporting the theory "that broker-sold, actively-managed funds underperform direct-sold, actively-managed funds because the direct-sold funds invest more in performance." 2 The funds sold through sales loads invest more in paying brokers for marketing.
There is an argument that this is the way things should be. People don't want to pay for financial advice, but some people need it, and if you're going to get them financial advice you need to find a way to hide the costs, and paying mutual-fund sales loads is a good way to make the costs easier to swallow. Sure, the costs are high and the performance is bad, but it's better than the alternative of not saving for retirement at all. There is probably something to this argument, 3 but it really does look unseemly to charge people extra for selling them a worse mutual fund.
This seems to be the Labor Department's focus: It thinks that mutual fund sales loads are bad and should be stopped, and it expects that the new rules will have that effect. So it says that the new rules will, "Lead to gains for retirement savers in excess of $40 billion over the next 10 years, even if one focuses on just one subset of transactions that have been the most studied," meaning front-end load mutual funds. 4 This is kind of low-hanging fruit, since front-end loads are already in decline: Less than 30 percent of mutual fund assets are in funds with front-end loads, and the number is declining by more than 2 percent a year. The Labor Department might be a bit behind the curve here, in that most investors have already figured this out, but I guess its job is really to protect the investors who are slow at figuring things out.
There's another specific focus in these rules, less pronounced but maybe more important than the thing about the mutual-fund sales loads. That is a bias toward passive investing. The proposed rules for the "Best Interest Contract Exemption" -- basically, you can charge all sorts of conflicted fees if you make a big song and dance about them -- include the possibility of a "low-fee streamlined exemption" that would exempt brokers from the song-and-dance requirements if they recommend the right products:
In this regard, the Department believes that certain high-quality investments are provided pursuant to fee structures in which the payments are sufficiently low that they do not present serious potential material conflicts of interest. In theory, a streamlined exemption with relatively few conditions could be constructed around such investments. Facilitating investments in such high-quality low-fee products would be consistent with the prevailing (though by no means universal) view in the academic literature that posits that the optimal investment strategy is often to buy and hold a diversified portfolio of assets calibrated to track the overall performance of financial markets. Under this view, for example, a long-term recommendation to buy and hold a low-priced (often passively managed) target date fund that is consistent with the investor’s future risk appetite trajectory is likely to be sound.
When we talked a while back about the White House report on conflicted retirement advice, I pointed out that most of the promised benefits to retirement savers came from some simple arithmetic in which (1) brokers currently steer savers to expensive actively managed funds, (2) brokers could instead steer the savers to cheap index funds, and (3) assumed pre-fee performance of the active funds and index funds is the same. I said at the time that the White House's announcement could almost have been titled "Strengthening Retirement Security by Cracking Down on Active Investing." The Labor Department proposals retain some of that flavor, though they are a bit gentler to active management.
I'm an index-fund guy myself, but I get a little nervous about the current vogue of regulatory love for passive investing, and regulatory hatred for investment management costs. We talked last week about the New York City pension funds, which pay fees for their public-market investments that are in line with what retail investors pay for index funds, though significantly higher than what big institutions pay for passive index investing. But the New York funds aren't invested passively: Their managers try to beat the market, and they do beat the market, and the funds get above-market returns even after subtracting the managers' fees. They don't beat the market by much, sure -- their outperformance after fees was about $40 million over 10 years, on a $160 billion portfolio -- but, hey, that's $40 million for New York City to spend on its retirees that it wouldn't have if it had just invested in the index at zero cost. 5
Even nonexperts can grasp a primal personal-finance principle: buy low-cost funds linked to the overall performance of the stock market, be patient and don’t try to outsmart the market or pay someone an arm and a leg to do it for you. That a succession of city comptrollers and fund trustees -- who, it should be noted, once included Mr. Stringer, a trustee in his old job as Manhattan borough president -- would never have thought of this before and found ways to reduce the damage done by excessive fees, is incredible.
But New York's pensions did better than they would have done by following that "primal personal-finance principle." Trivially better, sure, but it seems weird to conclude that New York City should have accepted worse performance -- and $40 million less for its retirees -- just to make a point about Wall Street greed and the joys of passive investing.
The Labor Department is far more balanced, but there's a lot more money in IRAs than there is in New York City's pensions, 7 so the Labor Department bestowing its official favor on passive investing would be a much bigger deal. We're still very far away from it, but reading this proposal I can almost imagine a future in which brokers to unsophisticated investors are required, or strongly encouraged, to sell them passive index funds, and active management is -- sort of like hedge funds and private equity now -- a semi-regulated domain of the rich and sophisticated-slash-foolish.
I've previously recommended this Josh Brown post from February about whether the decline of active management is cyclical or permanent. One theory that he cites, from GMO, is that active management outperforms in some environments and not in others, and that the current environment is not good for active management. But it will be again, and when it is, it will be awkward if millions of middle-class retirees have been pushed out of active investing by rules made today.
But the other theory, from Larry Swedroe, is that the decline of active management is more permanent, and
that the exodus into passive strategies -- which is usually cited as a reason for active outperformance going forward -- is actually a major negative. This is because, with all the unskilled investors departing, pros will be left to square off against only other pros. The lack of retail punters and their harvestable mistakes cuts off one of the most reliable historical sources of alpha for sharp-eyed managers.
Someone needs to allocate capital, and they need to be paid to do it, and historically they've been paid, indirectly, by schmoes who invest badly. 8 The Labor Department has its eye on one big group of schmoes: Underperforming active mutual funds sold through front-end loads, which invest in marketing through brokers rather than in outperforming the market through good investment decisions. If the new rules succeed in getting rid of those funds, it will be even harder for other active managers to outperform, and the bias against active management will continue to grow.
Moreover, in practice, disclosures of conflicts of interest can actually backfire. Research in behavioral economics and psychology finds that when advisers disclose their conflicts, they may be more willing to pursue their own interest over those of their clients and thus give worse advice. Investors may interpret the disclosure as a sign of honesty and become more likely to follow an advisers' biased advice.
Here's the paper, by Diane Del Guercio and Jonathan Reuter. Here's an NBER version, and a free one. The Labor Department's discussion (pages 119-120 of the RIA) is worth reading, citing Del Guercio and Reuter for the proposition that "actively-managed broker-sold funds track closer to indices than direct-sold funds" (because funds sold to investors are selling performance, while funds sold through brokers are not), and that broker-sold funds tend to have higher beta and less educated managers than direct-sold funds.
The paper itself, I should say, is a bit more charitable about motivations than the Labor Department is:
Evidence on the pricing and product characteristics of mutual funds sold through brokers versus those sold directly to investors supports our hypothesis that they serve investor clienteles with distinct preferences. Retail funds sold directly to investors offer unbundled access to portfolio management. Their investors neither receive, nor pay extra fees for, advice. In contrast, retail funds sold through brokers bundle portfolio management with financial advice. Del Guercio, Reuter, and Tkac (2010) show that fund families tend to sell their funds either directly to investors or through brokers, but rarely do both, suggesting a segmented market where the nature of competition differs across the two segments. Because experienced and knowledgeable investors are likely to self-select into direct-sold funds, flows in this segment are more likely to respond to risk-adjusted returns, giving direct-sold families a strong incentive to generate alpha. In contrast, the findings in Christoffersen, Evans, and Musto (2012) and Chalmers and Reuter (2012) suggest that competition in the broker-sold segment is likely to focus on characteristics other than alpha, such as the level of broker compensation. The weaker the sensitivity of investor flows to alpha, the weaker the incentive to generate alpha.
See the bit of Del Guercio and Reuter quoted in the previous footnote: This is kind of the standard model of load-based mutual funds.
From the RIA (page 8):
Focusing only on how load shares paid to brokers affect the size of loads IRA investors holding load funds pay and the returns they achieve, the Department estimates the proposal would deliver to IRA investors gains of between $40 billion and $44 billion over 10 years and between $88 and $100 billion over 20 years. These estimates assume that the rule will eliminate (rather than just reduce) underperformance associated with the practice of incentivizing broker recommendations through variable front-end-load sharing; if the rule’s effectiveness in this area is substantially below 100 percent, these estimates may overstate these particular gains to investors in the front-load mutual fund segment of the IRA market.
I exaggerate: That $40 million is basically nothing -- a few basis points, total, over 10 years -- and in fact index funds often beat their indexes by a little bit. So if New York had invested entirely in index funds it might have outperformed its benchmark by more than it did by investing in active funds. On the other hand it feels a bit like cheating to say that you should try to outperform the index by indexing.
Also I don't know anything about the risk/beta/Sharpe ratio/whatever of the actual funds, because the comptroller's report looked only at returns, albeit returns over 10 years.
Elsewhere, there's Annie Lowrey at New York Magazine ("Wall Street Is Fleecing New York City"):
As any financial adviser worth a damn would tell you, you would be better off investing in a low-cost index fund instead. Over time, you'd probably do about as well as the market does. And you wouldn't see all of your gains eaten up by fees.
But the pensions also do about as well as the market does (or very slightly better!), and also don't see all of their gains eaten up by fees. Or here is Hamilton Nolan at Gawker ("Oh My God Wall Street Is Robbing Us Blind And We Are Letting Them"), though I have omitted one word from this quote on this-is-a-family-blog grounds:
The public's pension money has grown by billions of dollars. How much of that will benefit the employees whose money it is? None! It all goes to the ... money managers!
The Times similarly said this, though it later corrected it:
If you take the funds’ gains since 2004, and subtract the fees, the analysis said, you end up basically at zero.
But these claims are simply false. If you take the funds' gains, and subtract the fees, and then subtract the gains, you end up at zero. But the gains after fees were more than 7 percent annually, or billions of dollars every year, all of which go to the employees. The fees ate up only (most of) the above-market returns. But if you just invest in the market, you don't have any above-market returns to eat up, so you are worse off than New York's pensions were.
The White House report cites $7 trillion of IRAs, of which "An estimated $1.7 trillion of IRA assets are invested in products that generally provide payments that generate conflicts of interest." The New York City pensions total about $160 billion.
Also directly by people who pay their management fees, but that still relies on the schmoes. If active management always and everywhere outperformed no one would pay for it; there needs to be some alpha to be gained by trading against uninformed investors.
To contact the author on this story:
Matt Levine at firstname.lastname@example.org
To contact the editor on this story:
Zara Kessler at email@example.com