I get a lot of questions about the mechanics of buying and selling the ETFs that comprise the tactical asset allocation (TAA) portfolios that I discuss on the blog. In this post I’ll address some common concerns and issues with respect to ETFs and share some tips for trading them in TAA portfolios.

When implementing buy and hold portfolios with ETFs the predominant concern is the management fee of the ETF, especially with an infrequent rebalancing approach. For example, with an annual rebalancing methodology at most there would be one trade per ETF per year. Other common trading concerns such as trading fees, liquidity (volume) and bid/ask spreads on the ETFs are less important in such an approach. Indeed, that is one of the strengths of the buy and hold approach. However, when implementing tactical asset allocation portfolios such as GTAA13 or GTAA AGG3, trading fees, liquidity, and bid/ask spreads take on a more important role due to the more frequent trading required by these strategies. Before we dive into the differences that TAA strategies may require in trading ETFs lets look at some good practices for buying and selling ETFs in general, regardless of the portfolio being used. Here are 5 tips I gleaned from this Vanguard white paper.

  1. Use “marketable” limit orders. I like this suggestion. Many investors get frustrated when using the common suggestion to use limit orders. For example, by trying to set a limit price at the middle of the bid and ask prices, investors could be waiting a long time for their order to execute. have the orders only partially execute, or have the orders not execute at all. The Vanguard suggestion is to use a limit order that is just above the ask price for a buy order and just below the bid price for a sell order. This will lead to faster execution and as long as the bid ask spread is not too wide us a reasonable approach.
  2. Beware of the time just after market open and just before market close. Usually, 5-min after the open is enough time to let the ETF prices ‘settle’ and the bid ask spread to stabilize. On the close you want to give yourself enough time to place the order and have the trade execute. Very near the close sometimes market on close orders can affect bid ask spreads quite a bit.
  3. Pay attention during volatile periods. There will be times where ETF prices, like any asset prices, will move quite rapidly causing bid ask spreads to widen considerable. It is usually best to let these times pass.
  4. Don’t pay too much attention to volume. This is a huge investor misconception with ETFs. For ETFs, the bid ask spread is a better indicator of liquidity than volume. See here for more. So, don’t worry too much about ETF volume.
  5. For international ETFs, consider the time of day. ETF prices represent the prices of the underlying stocks. If international markets are open during US trading hours the ETF can react to price changes in the international stocks. In general, morning represents a good time to trade international ETFs since most European markets are open until mid-day US eastern time.

Next, consider the details of implementing TAA portfolios. There are two key differences to pay attention to.

First, since the TAA systems require more trades, trading fees and the bid ask spreads impact returns more. Actually trading fees don’t matter that much for any reasonable size portfolio. Since most discount brokers offer flat trading fees regardless of volume and since many of the big ETFs (9 out of 13 in the GTAA 13 portfolio to be exact) are commission-free, trading fees are not nearly as important as bid ask spreads which are a percentage of the dollar value traded. Bid ask spreads change with market conditions but we can get an idea of average spreads by looking at the spreads over multiple trading days. Since most of the ETFs in the GTAA portfolios are Vanguard ETF lets take a look at their average bid ask spreads over the recent past. Vanguard keeps a running tally of bid ask spreads for their ETFs over the past month. See here. Pretty low spreads across the board. From 0.01% for some of the big asset class ETFs like VNQ to the less liquid asset classes like VGLT at 0.11%. I also looked up the average spreads of the ETFs in the GTAA13 portfolio and the recent average spread of the 13 ETFs is approximately 0.11%. Not bad. But it did highlight some outliers that I need to look into for potential replacement. IGOV, GSG, and IAU have the highest spreads among the 13 ETFs used. There may be better options available. Below is a snapshot.

GTAA13 avg bid ask spreads apr 2015

Second, the TAA system signals are based on the closing price of the previous day. Any trade that is done at a price that is different from the previous day’s close will affect returns. So, the goal is to execute any required trade at a price as close to the previous close as possible to minimize slippage. Good thing is that we can backtest these systems with buy prices executed at different times of the following trading day; the open price, the average of the day’s high and low, and the close price. Turns out that buying at the open gives the best results by 0.5% to 1% per year. The task then is to buy as close to the open price as possible when implementing these TAA portfolios. Personally, I implement all my TAA trades in the morning, usually 5-10 minutes after the open to let the bid ask prices settle, and use limit orders for all trades. If for some reason I cannot be on my computer the day I need to place trades then I will either use market on close (MOC) orders on the previous day or just use market on open orders on that day.

In summary, there are a few more details and issues to buying and selling ETFs when implementing TAA portfolios. With some care and attention to detail you can increase your chances of getting as close to to the ideal trade price as possible.

 


7 Comments

Tony Filly · April 15, 2015 at 8:03 am

Hi Paul,
Thank you so much for writing this blog… it’s awesome. I am definitely new to it and just getting rolling, so have a ton to learn…. but it would seem to me that waiting a day or a few hours would kind of be a wash over time… things might creep up or down a little bit and it would be hard to predict. I’m probably wrong, but would love to hear your comment.

    paul.novell@gmail.com · April 15, 2015 at 9:10 am

    Hi Tony,

    Well, no it’s not kind of a wash. Buying at the avg of the day’s hi and lo costs you 0.5% a year. Buying at the next day’s close costs you 1% a year. To get as close as possible to the theoretical result you need to buy as close to the open as possible. Now, even at 0.5% to 1% a year over the long term the system handily beats the market but then you need to adjust the theoretical returns to reflect this and then choose your system accordingly.

    Paul

John Klish · April 15, 2015 at 2:32 pm

Hi,

I’ve been a follower of yours and appreciate your work. With regards to any TIPS, I was wondering if you use anything to track your ETFs in your portfolio. Do you use excel or any other software? Would you mind sharing your thoughts? Thanks.

John

    paul.novell@gmail.com · April 16, 2015 at 9:27 am

    Hey John, I pretty much just use my brokerage account software (TD Ameritrade), and excel. For individual stocks I also use finviz.com.

    Paul

Karti · April 16, 2015 at 6:16 am

Hi Paul,

Really appreciate this blog post as I was struggling with the questions that you have addressed. This is very helpful and timely for me (along with the links that you have provided). Once again, thank you for this wonderful website and for sharing your knowledge and insights!

Karti

Damian · April 21, 2015 at 11:39 am

Paul, I appreciate all your contributions on the blog. Some other issues worth considering in the implementation of tactical portfolios (or, for that matter, any rebalancing approach) are these:

1. Fidelity, TD Ameritrade, and likely many other brokers require you to hold commission-free ETFs for at least 30 days before being able to sell them commission-free.

2. If you don’t have a margin account, many brokers make you wait to reallocate funds you receive from selling ETFs (or anything else) until the sale “settles,” which is several days later.

Both of these present annoying timing issues for tactical and rebalancing approaches.

With regard to the former, short months — and especially February — can be problematic. If you buy February 1, you won’t have held it for 30 days until March 3. Same issue is presented, albeit not to the same extent, by the 30-day months out there.

As for the latter, this is even more of a pain. If you sell your ETFs and then have to sit around for 3 or 4 business days for the sale to “settle,” you’re obviously not able to get in at the appropriate price for whatever you wanted to move into.

In cash accounts where you can get margin, you can alleviate this latter issue (albeit at a cost), but you can’t get around it in some accounts that won’t let you use margin.

Finally, just want to second your point about spreads. Some people I know are addicted to commission-free ETFs; in many cases, they are nice, especially if you’re dealing with smaller trades. You really have to watch those spreads, though. If you’re trading even a moderate amount — e.g., $3,000 — the spreads on some of the illiquid commission-free ETFs easily can outweigh the commission that you otherwise would have paid.

    paul.novell@gmail.com · April 21, 2015 at 12:33 pm

    Damian, great points to bring up. The trading fees don’t impact performance very much so I don’t think it’s worth the effort to get around point 1. Although it certainly doesn’t hurt to avoid the fees if possible. Hadn’t thought about the margin issue. I’ve always had a margin account so it’s never been an issue for me but important for those who can’t open a margin account.

    And the spread issue is huge. Bigger than all the rest. Especially when you add in the slippage you will already have between your execution price and the theoretical price the system has you buying at.

    Paul

Comments are closed.