(Gerard Michael, SAM) Brent Sullivan, author of the excellent The Tax Alpha Insider recently posted a summary of what he’s hearing from advisors about direct indexes. What emerges is that many advisors believe that $1mm - $5m is the “sweet spot” for direct indexing, that below $1m “the juice isn’t worth the squeeze” and that above $5m “Direct indexing is a gimmick”. (Brent isn’t necessarily endorsing these positions – he’s just reporting)
We respectfully disagree with this take, however common: our (apparently) contrarian view is that direct indexes make sense both for clients below $1mm and above $5m.
I wish to thank Brent for generously reviewing an earlier draft of this post and providing additional comments, which are included below. Any errors that remain are, of course, solely my responsibility.
Below $1m
Brent lists the top 5 reasons why advisors think direct indexing doesn’t make sense for investors with less than $1mm1:
Objection: Number of positions
Response: “Number of positions” is something of a catch-all objection, a variant of “meh, too complicated”. But if you find working with direct indexes to be fussy or time consuming, you’re doing it wrong. The management of direct indexes can be automated, making direct indexes as simple to work with as ETFs (if that’s not your experience, call us).
The number of positions is a barrier for smaller accounts if there are per-ticket commissions, but most major custodians now have $0 commissions, eliminating this concern.
Comment from Brent: “An adviser yesterday mentioned that transitioning individual stocks is a royal pain (vs. an ETF). Another adviser mentioned that “the number of line items on the performance report didn’t fit into a one-pager” and simply declined to consider individual stock positions any further.”
Objection: Tracking error
Response: The concern here is that small direct index accounts will suffer large tracking error due to the need to round each security holding to whole shares – in particular, the need to round down a lot of smaller positions weights to 0. Our experience is that for US large cap equities, this only starts being a problem for accounts below about $50K. Assuming a 50% allocation to US large cap, this means direct indexes can work for accounts as small as $100K. Once fractional shares become more widely available (admittedly, a prospect that forever seems just around the corner), this minimum drops to about $100.
You might ask why folks worry about tracking error in the first place. After all, it’s not something most investors care about. The real worry here is not tracking error, per se, it’s the risk that the direct index will miss out on a big rally by not owning some star performer – think NVidia, Tesla, Apple, etc. But in 20 years of working with direct indexes, we’ve never seen this be a problem in practice. Yes, there’s tracking error, but it’s usually lower than the documented taxes saved or deferred. And this sets up a “heads I win; tails I don’t lose” situation: if tracking error results in a positive return relative to the index, great; and if it results in negative return relative to the index, well, this negative return is less than the value of the tax management, so you’re still better off.
Comment from Brent: “This paper and its predecessor come up often when I hear these types of “omission error” arguments. The main point being that wealth is created by some tiny, unknowable, fraction of all stocks, so we better own every single one of them or else we’ll miss out. FOMO is real. Advisers will either argue the point directly to me, or they’ll say their clients care about this.”
Objection: Small/negative tax alpha
Response: Yes, the tax alpha for smaller accounts tends to be lower simply because the investor’s marginal tax rates tend to be lower. But the tax alpha is still quite high. We looked at accounts that were invested 100% in direct indexes. For accounts greater than $1mm, 100% cumulatively over the life of the account saved or deferred more in taxes from active tax management than the advisor charged in fees. Below $100K, this number drops to 78%. This shows that, indeed, direct indexing is more valuable for larger clients. But 78% is still an impressive number. And, to be clear, it isn’t that for 78% of accounts, taxes saved or deferred cumulatively made up for the management costs of the direct index. It’s that for 78% of accounts, taxes saved or deferred cumulatively made up for all fees paid to the advisor. The tax alpha is not small.
One twist here is that small-account holders with low marginal tax rates sometimes “graduate” into becoming large-account holders with high marginal tax rates. In this case, loss harvesting and other forms of gains deferral that are at the heart of tax management can be said to backfire – you’re effectively pushing gains taxes from a low to a high tax bracket. In practice, this is not a problem, even if it happens. It seems that most small investors are happy to exchange paying less tax when they’re poorer for more tax when they’re richer – the dollar value of the tax may be higher, but the pain it causes is lower.
Objection: Explaining and processing the portfolio
Response: As noted above, the management of direct indexes can be automated, so “processing” the portfolio should be no harder with direct indexes than with ETFs. On the other hand, explaining what’s going on may, indeed, take more time. Having more securities in an account does slightly increase the frequency of client calls. For large clients this is desirable. For smaller clients, maybe less so. And this is, by far, the main reason we see firms creating a size cutoff for direct indexing.
Comment from Brent: “The “explaining” part usually has to do with 1) the large number of line items on a tax return, 2) the large number of transitioning assets, and 3) what tracking error even is. What I’m hearing, understandably, is that clients don’t care to understand this stuff (who can blame them?), but are concerned when they’re paying more for a direct index SMA and getting hit by underperformance, versus an easy-to-understand ETF.”
{This may be a bit much, but additional comments from us in response to Brent’s additional comments: 1) as folks move to online statements and online tax filing, the “number of line items” concern basically goes away. 2) The number of transitioning assets is not an issue if you have automated the process. 3) No, no investor understands tracking error, but they do understand very high levels of documented tax savings or deferral; as for underperformance, tracking error is random, so ½ the time they’re outperforming the ETF, and, as noted above, the other half, they’re still ahead because of the tax benefits. The wonky direct index details of statistical sampling, tracking error etc can be ignored.}
Objection: Cost above a comparable ETF
Response: Competitively priced direct indexes are currently running in the 15 - 20 bps range (if you’re paying more, you’re probably paying too much). This is greater than the cost of comparable ETFs, but the price difference is dwarfed by the tax savings (see tax alpha discussion above). It’s true that the cost of direct indexing is more visible than the cost of ETFs (where the management fees are buried in the net asset value), but this issue with direct indexing is not confined to small accounts.
Above $5m
Brent proceeds to then list five reasons why direct indexes don’t make sense for investors above $5m2:
Comment from Brent: “The overarching theme is that while none of these opinions preclude direct indexing, many I speak with simply think there are either 1) lower hanging fruit on which to focus their limited effort/mindshare, or 2) better options for solving a given problem (e.g. long/short for single-stock concentration, rapidly depreciating assets for passive income offset, trader funds for mitigating ordinary income, etc.).”
Objection: Loss harvesting fades quickly
Response: Yes, if you simply buy a direct index and never add or subtract cash, the value of loss harvesting will fade. But this is not how portfolios are invested, so loss harvesting never actually disappears. More importantly, loss harvesting is not the main component of tax management. That honor belongs to risk-optimized gains deferral. And that does not fade. Every time you withdraw cash; every time you rebalance at the asset class level, direct indexes will enable you to reduce taxes relative to the alternative of having purchased an ETF. In practice, we do not see measured taxes saved or deferred declining over time.
Objection: Insufficient single-stock diversifier
Response: The concern here is that large portfolios often have to deal with highly concentrated positions, and wealth investors may want to take advantage of specialized solutions — like option collars, exchange funds, opportunity zones, charitable remainder trusts – for dealing with the problem. This is all true, but it doesn’t preclude using direct indexes as the primary source of market exposure.
Comment from Brent: “The case I hear often is of someone who has 80% of their wealth in a single name. In this instance, direct indexing cannot usually generate enough losses to effectively de-risk the position. Instead, investors opt for leverage (in the case of options or long/short), exchange funds, and charitable remainder trusts to get diversified fully (or close to it) quickly.”
Objection: More alpha available in private/real assets
Response: Maybe, but we’re skeptical. The same used to be said of hedge funds, but, when returns were analyzed, they underperformed simple index funds, partly (but not entirely) because something like 98% of returns went to managers. There are specialized vehicles – like gas, solar and real estate limited partnerships – that can be constructed to generate lots of tax deductions. And these may be good investments for wealthy investors, but, as noted above, this doesn’t preclude using direct indexes as the primary source of market exposure.
Objection: More tax alpha in estate planning
Response: Plausibly, but why not do both? Direct indexing and estate planning are not mutually exclusive. Even if estate planning is more important, direct indexes don’t take a lot of work or mind share, so why not use them?
Comment from Brent: “This is all about mind share, expertise and differentiation. If a tax planner is a specialist in multi-generation estate planning, as an example, they usually just focus on that to differentiate their practice since the estate tax is so visceral to UHNW. In other words, yes, they could do direct indexing, but many see solving the 40% estate tax as pure demonstration of competence, while direct indexing is a nice-to-have that won’t get many end clients too excited by comparison.”
Objection: Tax-aware long-short outperforms direct indexing
Response: This is another maybe. The basic idea is that if you invest in a roughly risk-neutral long-short position, you’re sort of guaranteed to generate loss harvesting opportunities no matter what: if the market goes down, the long position goes down (and can be loss harvested); if the market goes up, the short position goes down (and can be loss harvested). So you get more loss harvesting, and, in theory, you can keep this up for quite a while: in a rising market, the short positions, on average, start to wash out, but you can keep replenishing them as the total portfolio value rises. All this is great, but there are reasons to be cautious:
- Long/short direct indexes are more expensive than long-only direct indexes
- Shorting is usually expensive for individual investors, so much so that at some custodians it is simply economically impractical
- These strategies hold unhedged short positions. This carries unlimited risk. Yes, this doesn’t matter most of the time. But the extra risk is there, even if invisible. The added tax alpha may not compensate for this added risk.
- At present, long/short direct indexes are only offered as a separate product held in its own sub-account or sleeve, and, for both risk and tax efficiency, this is a bad way to manage direct indexes (See Direct Indexes Are Great. Direct Index SMAs Are Bad)
Comment from Brent: “The extra nuance I’d add is that long/short is a solution to a particular problem, namely a lot of consistently-realized gains somewhere else in the portfolio. So if an investor has that particular problem, then the risk from unhedged shorts could make sense. And many who manage shorts usually limit their exposure to large cap, where the borrowing capacity is nearly unlimited. It’s just another tool in the toolbelt.”
It should be noted that Brent is not endorsing the various objections listed. He’s just acting as a faithful reporter, relaying what real-life advisors are thinking, right or wrong. And he notes: “many advisers expressing concern about the operational headaches of direct indexing, including the number of tax return line items and transitioning assets, etc., have not adopted direct indexing. In other words, many are expecting pain, not actually experiencing it.”
Brent ends his post with the questions “What do you think?”. It’s probably a good way to end any post. We’d love to hear your thoughts.
1. Two reasons Brent lists that advisors give for using direct indexing for accounts below $1m: managing concentration and employer risk, and customizing portfolios for impact.
2. Two reasons Brent lists that advisors give for using direct indexes for accounts over $5m: investor (and advisor) risk tolerance, and advisor access (e.g. the right custodian and platform/network).
