Advisor Perspectives welcomes guest contributions. The views presented here do not necessarily represent those of Advisor Perspectives.
Advisors who run or plan to run ETF-based portfolios need to have a formalized trading methodology. For those who haven’t yet developed one, this article is intended to help accelerate progress and avoid some risks that may not be obvious to anyone who is primarily experienced with trading mutual funds.
I won’t offer reasons to switch to ETFs. There are plenty of resources to help with that decision. But the mass migration from mutual funds to ETFs has progressed farther than anyone anticipated when they first launched in 1993, and the recent approval of Dimensional Fund Advisors’ application for ETF share classes — which likely implies the opening of the floodgates for such structures — may only serve to accelerate this trend. My goal in this article is to share the results of my advisory firm’s research, in the hope it might make a few advisors’ migrations a bit smoother.
Important and hopefully obvious disclaimer: Developing a first-class trading policy is each RIA’s own responsibility. While my goal with this article is to share what I’ve learned in the hopes of lowering some hurdles to developing a top-notch policy, I cannot and do not bear any responsibility for any outcomes.
It's Not the Same Old Thing
The first step, as they say, is to admit you have a problem. Mutual fund trading policies that have worked just fine for you will no longer serve. ETFs are a different beast. Not hugely so, but different enough to cause problems if you don’t understand the differences. Best execution is not as simple with exchange-traded vehicles. Below, I outline some of the reasons why, and what can be done to address them.
Beware Drift in Transition
If you sell a mutual fund portfolio to buy an ETF portfolio, the purchases will happen before the sales, even if you enter the sell orders first. Mutual fund orders are executed after the market closes, whereas ETF orders are essentially stock trades, executed during the market day.
The risk is that the ETF purchases will be too large for the sale proceeds, due to market drift in the interim. The problem may be exacerbated by rebalancing software, which typically relies on last-best-known pricing data, meaning the tools will estimate mutual fund sale proceeds based on prior-day closing prices while calculating ETF purchase orders using recent, same-day pricing. On a down day for the markets,1 this may mean proposed buy orders will be derived from sell orders that are expected to fetch more than they really will.
The most conservative solution to this problem is to wait to buy the ETF portfolio until the trading day following the liquidation of the mutual fund portfolio. But sitting in cash overnight is suboptimal. Less stringently conservative strategies include:
- Wait until late in the day to execute your trades, to minimize the risk of market drift after the new portfolio is purchased.2
- If markets have moved substantially intraday (especially downward), consider holding off a day on the transition.
- Especially if markets have been volatile, set aside a few percentage points’ worth of cash in transition, large enough to ensure that even a wild end-of-day market drop won’t lead to being overallocated. You can top off the portfolio by making additional purchases the next day.
Don’t be Freddie the Freerider3
A “freeride” violation occurs when you buy a security and then sell it again before the cash required to buy it in the first place has settled in the account. Ignore Edgar Winter Group’s recommendation to come on and take a free ride, lest your clients’ accounts be subjected to trading restrictions.
Because mutual fund trades execute after the market close and settle before the next open, a freeride violation on a mutual fund portfolio is so improbable as to be almost a braggable accomplishment.4 Thankfully, with the move from T+2 to T+1 settlement for equities (including ETFs), freeride violations are also less probable with ETFs.
The biggest risk may come from realizing too large a purchase was made and trying to reverse it with a same-day sale. This doesn’t work! Instead, a partial reversal should be made through your custodian’s trade error process. Each custodian handles this differently, so this will require some research.
Block ‘em Up!
When you’re running mutual fund portfolios, each trade is independent. When you’re trading ETFs, though, it’s preferable to do as few independent trades in the same ticker on the same side (buy or sell) as possible. This is because average pricing tends to degrade when you execute multiple trades in the market, and because doing so results in different prices for different client accounts, introducing undesirable randomness into account-by-account trade execution.
The solution is to “block up” your trades. Same ticker/same side transactions can be combined into a single block trade. The trade will be executed in your firm’s trading account and then distributed back into the client accounts on which the trades were originally proposed. Here again, the exact method for executing and allocating block trades will vary, not only by custodian but by the specific tool you are using for trade execution. The research effort is worth it!
One element to be aware of: Weird things5 can occur if you execute a block trade but then forget whatever final steps are necessary to allocate the trade back into client accounts. It’s worth explicitly building each step of the block trading and allocation process into your trading rules and trading habits.
Avoid the Auctions
Crazy stuff6 can happen with equity prices near the “opening auction” and “closing auction” at either end of the trading day. Conservative advice I’ve heard from trading desks is to avoid the first and last half-hour of the market day, if possible. Holding off on placing trades at other volatile times, such as the general vicinity of Fed decisions or major economic releases, is also a good idea. Exigencies being what they are though, like the pirates’ code, this recommendation is more of a guideline than an actual rule.
Don’t Push an Elephant Down a Straw
If I had to pick a big one, this would be it. Though I’ve thankfully never experienced it, I have heard horror stories about advisors submitting multi-million-dollar ETF trades as market orders and winding up with average execution multiple percentage points outside the bid/ask spread. That’s tens of thousands of dollars of pain from a single trade execution mistake.
When you buy and sell mutual funds, figuring out how to raise or deploy the cash most efficiently is the fund manager’s problem. You simply get the closing NAV. Unfortunately, this means inefficiencies (cash drag, etc.) deriving from every investor’s trading activity are shared proportionally by all investors in the fund.7
With ETFs, the execution costs of moving in and out of the fund are generally borne by the investor making the trades instead. Consequently, you typically won’t be hurt by the actions of other investors or advisors who trade unintelligently.8 But if you trade unintelligently…
Even ETFs that trade very little (on the secondary market) can source liquidity from their underlying asset base (the primary market). But this feature is not automatically activated if you press go on a market order. Instead, market orders are typically first offered to market-making high-frequency traders. But market makers may refuse large trades,9 which then risk “sweeping the order book” — meaning some folks who’ve placed limit orders far outside the current bid/ask spread benefit at your clients’ expense.
The goal of each trade should be to land at or inside the National Best Bid/Offer (NBBO), which is to say that you want every ETF trade to do no worse than hitting the highest bid price (for sell orders) or lowest ask price (for buy orders) quoted on national exchanges at the time the order is executed. If trades are small, this is typically easily accomplished in a market order, as market makers earn “deal flow” — i.e., they earn the right to be given trade orders — largely by executing small trades at good prices. But if an order is sizable, a different approach is required.
Lesser superpower: Executable limit orders. To avoid the risk of getting a trade executed far outside the NBBO, one option is to place executable limit orders. For buy orders, this means setting a limit a penny or two above the currently quoted ask; for sell orders, set the limit a penny or two below the bid. Most orders set with these precautions will execute just fine. Occasionally, the market may move against you, and you’ll have to shift the limit to account for the new bid/ask range. But the risk of massive slippage will disappear.
The RIAs I work for require executable limits on 1,000- to 2,000-share orders. However, we rarely see a trade with negative price improvement (i.e., priced outside the NBBO), which suggests this precaution may be overwrought.
Greater superpower: The block desk. Every custodian employs trading specialists in a team called a “block desk.” The block desk is your go-to solution for placing large orders (which, despite the name, may or may not be block orders) at good prices. How large is “large”? You should seek out your own information on this, but I have had multiple custodians tell me the magic number is 2,000 shares.
This seems odd, given that the dollar amount represented by 2,000 shares can vary wildly. And popular, heavily-traded ETFs could surely handle a higher load than could low-volume ETFs. But 2,000 shares is a simple rule to follow and seems to work well.
You route an order to the block desk by marking it “not held” in the order details. So held means “kick this out the door immediately” while not held means “hold on to this until we chat and/or until you can work out a good price.” Whatever. I don’t make the jargon.
Talk with your custodian about how to navigate block desk mechanics. Some desks may assume you want to execute an order “best way” (i.e., the best price they can get when they receive the order) unless you indicate otherwise in the order details. Others may want to call you for instructions, in which case best execution/best way is usually what you want, though it can be comforting to wait and learn over the phone what price that implies.
The block desk is empowered to work with exchanges, market makers, fund managers, and authorized participants10 to find the best available price. I’ve never experienced a problem11 with block desk-executed trades, even on multi-million-dollar orders for ETFs with low volume.
Finally, as was already true with mutual funds, if you plan to place large trades, it can be helpful to communicate with the fund manager beforehand, to let them know the trades are coming and to ask if they have any additional requests or instructions. I recommend seeking the fund manager’s guidance on what “large” means in this context, but it’s likely millions of dollars.
You Can’t Manage What You Can’t Measure
Custodians are willing to send periodic (typically quarterly) reports on execution quality for your firm’s trading activity. Take them up on this! If you can’t see how you’re doing, you won’t know what you could improve!
Tap Your Custodians’ Expertise
As you build out your company’s ETF trading policies, I strongly recommend leaning on your custodian’s expertise. Don’t be afraid to ask questions or to navigate the bureaucracy to get your questions answered by the right people.
Don’t be intimidated by the risks. You’ve got this! A well-designed trading policy will merely add to the list of reasons why you’ve chosen to move to ETFs in your advisory practice.
Endnotes
1 While this risk is most acute for stocks and other volatile assets, “markets” here refers to whatever underlying asset mix inhabits your model portfolio.
2 This also minimizes the amount of time effectively spent in a 2X-levered portfolio — holding both the ETFs and the mutual funds — the mirror image of spending too much time in cash.
3 What’s an investment article without an obscure Red Skelton and/or Miles Davis reference?
4 Kidding!
5 Technical term.
6 Also a technical term.
7 As an aside, many advisory firms have rules requiring that their client base remain below a certain threshold percentage of a mutual fund’s AUM. But if you are good about keeping the fund manager in the loop — so that they can coordinate trading activity in anticipation of any large trades you need to place — this makes little sense. The real concern is that some other advisor is a large percentage of a fund where your clients are a smaller percentage, because then the other advisor’s failure to communicate can cause problems for your clients, given the sharing of internal trading costs across the full pool of investors.
8 An unfortunate potential side effect for ETF share classes is that, because the ETF and the mutual fund share a common pool of assets, poor investor behavior on the mutual fund side may still affect the returns of the ETF.
9 I’ve heard that market makers themselves will often be given significantly worse execution on large trades as well, perhaps because they are at least partly judged on a percentage price improvement per trade basis, rather than exclusively on price improvement per dollar traded.
10 The authorized participant is the entity capable of creating new shares with baskets of underlying securities and redeeming existing shares by releasing the underlying securities back into the wild. This entity may or may not be the primary market maker for the ETF as well.
11 I have seen block orders land very slightly outside the NBBO, with slippage on the order of a few dollars on trades of hundreds of thousands to millions of dollars in size. But even this is rare; block desks can get price improvement far more often, and most trades through the block desk will at least hit the bid or ask.
In his roles as chief investment officer for Round Table Investment Strategies and portfolio manager for Torren Management, Nathan Dutzmann is responsible for applying financial science and investment research to the process of constructing portfolios tailored to the individual needs and goals of clients nationwide. Nathan was previously an investment strategist with Dimensional Fund Advisors and a partner and chief investment officer with Aspen Partners. He is also a member of the investment industry advisory council for The American College of Financial Services. He holds an MBA from Harvard Business School and a master’s degree in international political economy and a bachelor’s degree in mathematical and computer sciences from the Colorado School of Mines.
A message from Advisor Perspectives and VettaFi: Discover something new! Click here to register for our upcoming webcasts.
Read more articles by Nathan Dutzmann