SMAs are better than mutual funds for most assets. But they won’t replace mutual funds entirely.
Mutual funds were invented in the 1920s as means to reduce transaction costs enough for ordinary investors to own a diversified portfolio. Trading costs ー bid/ask spreads and commissions ー were something on the order of 100 times more expensive than what they are today, making separately managed accounts (SMAs) impractical for all but the very wealthy.
That’s changed. Today, mutual funds are more expensive, less tax-efficient and less customizable than a well-run models-based separately managed account, at least for US Large Cap equities.1 Does that mean mutual funds are headed for extinction? We don’t think so, and we’d like to share why.
The Case for SMAs
First, let’s summarize the case for SMAs:
SMAs are substantially less expensive than actively managed mutual funds. The average fee for actively managed funds is roughly 80 bps. Models-based SMAs cost roughly 45 bps ー almost half. (There are overpriced SMA programs, of course, but we’ll ignore those.)
SMAs are more tax efficient than mutual funds, to the tune of 50+ bps in increased after-tax returns. Mutual funds suffer from three tax-inefficiencies:
- Mutual funds can generate capital gains distributions solely as the result of other fund investors selling shares, meaning that even buy and hold investors are subject to capital gains taxes.
- Mutual funds are barred from passing net tax-loss harvesting benefits to investors.
- When selling a mutual fund, owners cannot reach into the securities underlying the mutual funds and preferentially sell those with the least gain. They can with SMAs. That is, if you own a mutual fund that went up in value by 10%, you’ll realize 10% in capital gains for every dollar sold. There’s no way around it. But if you directly own the underlying shares in an SMA basket of stocks that went up by 10%, on average, you have the option to sell the least appreciated positions ー possibly selling only positions with losses.
Suppose you work for Exxon, hold Exxon stock options, and think that you shouldn’t have any further exposure to Exxon stock? You can’t ask a mutual fund to “hold the Exxon.” You can with an SMA. The same applies for social criteria constraints, sector constraints etc.
Equal rebalancing simplicity
Mutual fund portfolios used to be easier to rebalance than SMAs. But with modern rebalancing software (like ours), that is no longer true. Rebalancing a portfolio with individual securities is no more operationally complex than rebalancing a simple mutual fund or ETF portfolio.
The Case For Mutual Funds
With all of these SMA advantages, why does anyone invest in mutual funds? We see five reasons:
Good for small accounts
Mutual funds scale down to accounts of any size. SMAs don’t. The need to own whole shares and/or pay per-ticket commissions means that SMAs don’t make sense below roughly $50K per asset class. (Some custodians, like Apex and FOLIOfn, support fractional shares and don’t charge a per-ticket commission. This allows SMAs to compete effectively with mutual funds on price for portfolios as small $50 per asset class.)
Conceptual simplicity ー not a lot to explain to clients
With SMAs, clients will own individual securities. This can invite unproductive conversations about specific companies and their performance. It also increases the likelihood of extreme performance outliers. Some advisors prefer the simplicity of showing clients fewer holdings.
Good for illiquid asset classes
The mutual fund structure comes into its own for securities ー like corporate bonds, foreign stocks and foreign bonds ー that are not easy to trade. For these securities, the mutual fund manager is able to add value through their trading expertise. Model-based SMAs work best with securities, like US large cap equities, that are easy to trade.2 For this reason, it’s rare to see accounts that use SMAs exclusively. It’s more typical to see SMAs for core asset classes and mutual funds for peripheral asset classes. (This mixture of SMAs and mutual funds is sometimes called a Unified Managed Account, or UMA. See our previous post A Guide to SMAs and UMAs.)
Fees are hidden
To a first approximation, mutual fund fees are invisible. They're taken right out of the fund and show up only in the form of a lower price for the mutual fund shares. The fees are disclosed in the fund prospectus, but few investors read these. Not so with SMAs ー their fees show up as prominent fee deductions on a client’s statement. This creates the false impression that SMAs are more expensive.
We’ve had clients who set out to convert their mutual fund clients to UMAs precisely because UMAs were better and less expensive. With few exceptions, these conversions haven’t been particularly successful. The problem is that it takes effort to explain the change to clients, and it’s very hard to overcome the hidden fee problem. The client is told their fees are going down, but at the same time a large new set of fees shows up on their statements ー which can make the client feel as though they’re worse off. In the end, most firms seem to prefer to leave well enough alone. They may start switching new clients to UMAs, but they mostly leave legacy mutual funds clients where they were.
No new systems
We stated in the beginning that SMAs are less costly than mutual funds, yet more customizable and tax efficient. This is only true for firms that have access to modern rebalancing/implementation systems. Not all firms do, and for firms that don’t, mutual funds remain a superior (or, more likely, the only) option.
Bottom Line: Mutual Funds or SMAs?
So which is better, mutual funds or SMAs? Here’s the summary:
SMAs are better for liquid asset classes, like US equities, for accounts with $50K or more to invest in each asset class. With a new generation of custodians that support fractional shares and don’t charge per-ticket transaction cost, SMAs are better for liquid asset classes more or less across the board.
Mutual funds are better for assets where trading expertise is central to the value third-party managers provide. They’re also better for smaller accounts (barring, as noted above, the option to invest in fractional shares and not pay per-ticket commissions).
The greatest disadvantage of SMAs relative to mutual funds is that SMA fees are prominently disclosed, whereas mutual fund fees are “hidden” in the fund’s share price. We think this is the primary reason the outflow from mutual funds hasn’t been greater, but we don’t think it’s a permanent barrier. Newer proposal systems are starting to improve fee transparency, “uncovering” the mutual fund fees in a way that makes them easy for investors to see. Firms that have large existing mutual fund books may be reluctant to embrace this, but newer and more agile firms will not.
The advantages of SMAs do require advanced rebalancing/implementation technology (yes, like ours). And that is obviously a barrier to adoption, but it’s not permanent. The ability to manage customized, tax-optimized SMAs at scale is becoming table stakes.
So, mutual funds or SMAs? We expect that the answer is both.
1 There are two types of SMAs. In a models-based program, one or more asset management specialists provide a model portfolio (a data file) to an overlay manager, who is responsible for all customization and tax management. In a traditional SMA, each asset management specialist directly controls and trades a sub-account. In this post, we only talk about models-based SMAS. Traditional SMAs have only modest advantages over mutual fund ー they are, in fact, sometimes criticized as “mutual funds in disguise.”
2 In principle, it is possible to implement models-based SMAs for illiquid security types. The difference is that the asset management specialists needs to provide trade recommendations by security characteristic (e.g. bond duration, quality, etc.) instead of providing specific security identifiers.