ETF Analysis: Do Historical Returns Predict Future Returns?

Article Purpose

Historical results are among the first things many investors consider when analyzing ETFs. Except for contrarians, investors search for funds with solid one-year, three-year, and five-year returns, as this is evidence the strategy can work, notwithstanding macroeconomic changes or changes in the portfolio’s composition. But you might be wondering how well this simplistic strategy works and determining that is the purpose of today’s article. In this report, I will evaluate the “performance-chasing” strategy by looking at trailing and forward one-year, three-year, and five-year returns by segment and style for hundreds of U.S. Equity ETFs. I think you’ll find the results helpful, and I hope you can use the information in this article to become a better investor.

ETF Performance Analysis

Background

Many readers are familiar with Portfolio Visualizer, a free tool that lets you compare the risk and returns of up to four securities at a time. You can also “optimize” your portfolio by entering a set of assets and the period to evaluate, but there’s one catch. By default, the tool uses an asset’s historical metrics as inputs. As a result, the output is what asset combination would have worked best in the past, thereby providing some investors with a false sense of security. While it appears they’re doing proper due diligence, they’re just checking past performance.

One benefit of this article is that I’ve done a lot of this backtesting for you. After gathering historical returns for hundreds of U.S. Equity ETFs, I intend to use the statistics to answer two key questions:

1. Are one-year, three-year, and five-year trailing returns good predictors of an ETF’s success, as measured by their corresponding one-year, three-year, and five-year forward returns?

2. Are the results similar depending on the segment and style of the market (small-cap, mid-cap, and large-cap value, blend, and growth)?

This analysis is lengthy, but I’ve tried to organize the results in such a way that makes it a relatively easy read for you. Hopefully I’ve done my job, but let’s get started by trying to answer the first question.

Trailing Returns By Segment

There are numerous ways to evaluate and rank performance. However, the most simplistic way is to look at all ETFs irrespective of the segment (small, mid, and large) and style (value, blend, and growth). If an investor bought the top-performing ETFs on December 31 and held them for a period that matches the evaluation period, would they come out ahead?

For the most part, the answer is no, but it depends on the segment. Using over 4,000 data points for 576 U.S. Equity ETFs over 23 years, investors selecting the top 25% of ETFs based on their trailing one-year returns ended up with slightly below-average returns the following year (55th percentile).

• Quartile 1: 55%
• Quartile 2: 52%
• Quartile 3: 49%
• Quartile 4: 48%

These results suggest contrarian investors might be onto something. ETFs ranked in Quartiles 3 and 4 had above-average performance the following year, though the difference is slight. However, I also calculated figures based on segment, as comparing two ETFs in different categories (e.g., small-cap vs. large-cap) is uncommon. Here, we see some performance differentiation, summarized in the table below.

The contrarian approach worked better with the small-cap and mid-cap segments, but I’m still not confident one-year trailing returns are anything more than an interesting data point. Therefore, I recommend tossing this statistic completely.

The results are no better using three-year trailing returns, though the performance-chasing strategy worked better with large caps. The top two quartiles did slightly better (49th percentile) for the next three years.

Lastly, the following table indicates that large-cap investors buying ETFs based on their five-year trailing returns have a meaningful advantage, with the top quartile ranking above average (44%) over the next five years on average. Still, the opposite was true for small-cap and mid-cap ETFs.

Here are three takeaways:

1. Analysts should not use one-year and three-year trailing returns to justify an investment in an ETF. Similarly, they should not discount an ETF because of its poor one-year and three-year track records. They are fun data points but should not dominate an analyst’s investment thesis.

2. Five-year trailing returns are better predictors of forward five-year returns for large-cap ETFs. ETFs ranked in the top two quartiles performed above average over the next five years. While it’s not a slam-dunk strategy, this approach earns you a slight advantage and can help narrow your choices.

3. There is no evidence that the best-performing small-cap and mid-cap ETFs over the last one, three, and five years will perform well moving forward. The reason could be because less is known about the stocks held in these funds or because the sample size isn’t yet large enough. While my sample size for large-cap ETFs over five years is sufficient at 108, it’s only 43 and 34 for small-cap and mid-cap ETFs.

Trailing Returns By Segment and Style: Large-Cap ETFs

We can further divide the analysis by style (value, blend, and growth), but due to the limited information available for small-caps and mid-caps, I’m only comfortable doing this with large-caps. I’ve summarized these results below.

Using three- and five-year trailing returns, large-cap growth ETFs ranking in the top quartile went on to rank in the 41st and 46th percentiles in their category over the next three and five years. Interestingly, the top-quartile performers in the large-cap value and large-cap blend categories ended up ranking below average on all periods measured.

Large-Cap Value ETFs: In-Depth Analysis

As always, there are exceptions worth noting. Consider the following table of historical and forward five-year returns for the 30 large-cap value ETFs in my sample as of December 31, 2016.

These results speak for themselves and demonstrate there is no correlation between historical and future five-year returns for large-cap value ETFs for the year ending December 2016. Instead, what might be more helpful is examining above-average performers in two distinct five-year periods (i.e., 2012 to 2016 and 2017 to 2021). This analysis would prove consistent performance over ten years and would narrow your choices to the following:

1. BrandywineGLOBAL Dynamic U.S. Large Cap Value ETF (DVAL)
2. Vanguard Value ETF (VTV)
3. Vanguard Mega Cap Value ETF (MGV)
4. iShares Core S&P U.S. Value ETF (IUSV)
5. SPDR S&P Dividend ETF (SDY)
6. iShares S&P 500 Value ETF (IVE)
7. SPDR S&P 500 Value ETF (SPYV)
8. Vanguard S&P 500 Value ETF (VOOV)

If you’re not looking to shoot for the moon, it’s a sensible approach. You’ve already decided you want exposure to this market corner and simply wish to increase your chances of success before digging into an ETF’s strategy and fundamentals, which can be time-consuming. Next, let’s fast-forward one year to December 31, 2017. Large-cap value investors on this date would have 35 choices, 30 of which I’ve listed below.

Again, there isn’t a good correlation between historical and forward five-year returns. However, if we look for the above-average performers (i.e., those ranked #1-17 in both five-year periods), we can quickly narrow things down:

1. BrandywineGLOBAL Dynamic U.S. Large Cap Value ETF (DVAL)
2. Schwab U.S. Dividend Equity ETF (SCHD)
3. SPDR S&P Dividend ETF (SDY)
4. Vanguard Value ETF (VTV)
5. Vanguard Mega Cap Value ETF (MGV)
6. iShares Select Dividend ETF (DVY)
7. SPDR S&P 1500 Value Tilt ETF (VLU)
8. Vanguard High Dividend Yield ETF (VYM)
9. First Trust Value Line Dividend Index Fund (FVD)
10. WisdomTree U.S. Total Dividend Fund (DTD)

Finally, let’s jump forward one more year to December 31, 2018. On this date, there were at least 39 large-cap value ETFs to choose from, and the Invesco S&P 500 High Dividend Low Volatility ETF (SPHD) was at the top of the list with a 62.15% total return from January 2014 to December 2018, confirmed here. However, SPHD would become the worst performer in its category over the next five years.

From 2014-2023, the most consistent large-cap value ETFs were:

1. WisdomTree US Quality Dividend Growth ETF (DGRW)
2. ProShares S&P 500 Dividend Aristocrats ETF (NOBL)
3. Schwab U.S. Dividend Equity ETF (SCHD)
4. Vanguard Mega Cap Value ETF (MGV)
5. Vanguard Value ETF (VTV)
6. WisdomTree U.S. LargeCap Dividend Fund (DLN)
7. Vanguard Dividend Appreciation ETF (VIG)
8. WisdomTree U.S. Total Dividend Fund (DTD)
9. FlexShares Quality Dividend Index Fund (QDF)
10. SPDR S&P 1500 Value Tilt ETF (VLU)

Large-Cap Blend ETFs: In-Depth Analysis

Next, let’s move on to large-cap blend ETFs, starting with what the picture looked like on December 31, 2016. As shown, there were very few instances where the top performers over the last five years were also the top performers over the next five years.

There are 41 ETFs in my sample, so I consider those ranking in the top 20 over both five-year periods to be “consistent” ETFs. They are as follows:

1. Invesco S&P 500 High Beta ETF (SPHB)
2. Vanguard S&P 500 ETF (VOO)
3. iShares Core S&P 500 ETF (IVV)
4. Vanguard Total Stock Market ETF (VTI)
5. iShares Russell 1000 ETF (IWB)
6. Schwab U.S. Broad Market ETF (SCHB)
7. SPDR S&P 500 ETF (SPY)
8. SPDR S&P 500 Composite Stock Market ETF (SPTM)
9. Vanguard Large Cap ETF (VV)

Interestingly, except for SPHB, all are plain vanilla funds with a simple market-cap-weighting scheme. ETFs with alternative weighting schemes ranked #22/41 on average from 2012 to 2016 and #28/41 from 2017 to 2018. Therefore, the evidence over these ten years indicates that when it comes to straightforward large-cap blend exposure, the easiest solution is probably the best solution.

One year later, on December 31, 2017, investors had at least 44 large-cap blend ETFs with five-year histories. The First Trust Capital Strength ETF (FTCS) and the Invesco S&P 100 Equal Weight ETF (EQWL) were good choices for the next five years, ranking #15/44 and #9/44, respectively, from 2018 to 2022. In this sense, the performance-chasing strategy worked well.

However, raise your hand if you bought the VanEck Morningstar Wide Moat ETF (MOAT) based on its 103.98% total return from January 2013 to December 2017, which ranked just #30/44. It went on to rank #1 with a 63.85% total return over the next five years, and this recent success led VanEck to create more “Wide Moat” ETFs like MGRO, MVAL, MOTE, and SMOT. Granted, the five-year trailing returns for all 30 funds listed above are pretty close, so while MOAT didn’t stand out then, it shouldn’t have been ignored just because it lagged behind SPY by a few percentage points.

Still, following the same method as earlier for determining the most consistent performers, there were 15, as follows:

1. First Trust Capital Strength ETF (FTCS)
2. Invesco S&P 100 Equal Weight ETF (EQWL)
3. Invesco S&P Quality ETF (SPHQ)
4. iShares Russell Top 200 ETF (IWL)
5. Invesco S&P 500 Revenue ETF (RWL)
6. Vanguard Mega Cap ETF (MGC)
7. Vanguard S&P 500 ETF (VOO)
8. iShares Core S&P 500 ETF (IVV)
9. SPDR S&P 500 ETF (SPLG)
10. Vanguard Large Cap ETF (VV)
11. SPDR S&P 500 ETF (SPY)
12. SPDR S&P 1500 Momentum Tilt ETF (MMTM)
13. Schwab U.S. Large-Cap ETF (SCHX)
14. SPDR S&P 500 Composite Stock Market ETF (SPTM)
15. Vanguard Russell 1000 ETF (VONE)

This group has some funds with equal-weighting schemes (FTCS, EQWL), but 10/15 follow plain vanilla market-cap-weighted strategies. SPHQ also did well, but as discussed in this article, the highest-quality companies tend to be the largest. Underlying profitability metrics like weighted average net income margin and return on total capital figures for the plain vanilla ETFs listed above strongly support my claim that they are high-quality ETFs, too.

Let’s skip ahead one more year to December 2018, when there were at least 50 large-cap blend ETFs with a five-year history. Interestingly, the two ETFs at the top of the list followed different strategies (momentum and low-volatility) and outpaced most plain vanilla funds by a wide margin. Unfortunately, investors counting on this performance to repeat were disappointed. MTUM and USMV ranked #46/50 and #47/50 over the following five years, with DIA, LGLV, SPLV, and FTCS ranking in the bottom 20%.

After we get past the top six ETFs mentioned earlier, the forward five-year returns for the top performers end up pretty strong. The most consistent ones were as follows:

1. VanEck Morningstar Wide Moat ETF (MOAT)
2. Invesco S&P 500 Top 50 ETF (XLG)
3. iShares Russell Top 200 ETF (IWL)
4. Vanguard Mega Cap ETF (MGC)
5. iShares Core S&P 500 ETF (IVV)
6. Vanguard S&P 500 ETF (VOO)
7. iShares S&P 100 ETF (OEF)
8. SPDR S&P 500 ETF (SPY)
9. Invesco S&P 500 Quality ETF (SPHQ)
10. Vanguard Large Cap ETF (VV)
11. Schwab U.S. Large Cap ETF (SCHX)
12. SPDR S&P 500 ETF (SPLG)
13. iShares Core Total U.S. Stock Market ETF (ITOT)
14. iShares MSCI USA Quality Factor ETF (QUAL)
15. iShares Russell 1000 ETF (IWB)
16. Vanguard Russell 1000 ETF (VONE)

Large-Cap Growth ETFs: In-Depth Analysis

Lastly, let’s look at large-cap growth ETFs, a much smaller group. As of December 2016, only 18 funds had a five-year history.

As shown, the Invesco QQQ ETF (QQQ) led by a long shot from January 2012 to December 2016, delivering a 125.46% total return. QQQ also ranked #1 from January 2017 to December 2021, with a 248.19% total return, and for many investors, the case is closed. However, consider that applying the same logic didn’t work out so well with RPV and BMVP for the same 2012-2016 period or for investors clinging to low-volatility ETFs like SPHD, USMV, LGLV, and SPLV in December 2018. Unfortunately, the only sure thing is that more analysis is required.

Fast forward to December 2017, and the picture looks similar. QQQ is still out in front, though its five-year forward ranking dropped to #2 behind the Invesco S&P 500 GARP ETF (SPGP), the worst-performing fund from 2012 to 2016. For those unfamiliar with SPGP, it changed Indexes in June 2019, so I place little emphasis on its historical results. However, Index changes can render past performance virtually useless, and they are common for those holding Invesco ETFs. To be sure, you should always read the footnotes in the performance section on the fund page.

QQQ has been consistent, but so has the iShares Russell Top 200 Growth ETF (IWY), the Vanguard Russell 1000 Growth ETF (VONG), and the iShares Russell 1000 Growth ETF (IWF). Compared to QQQ, they’ll never look good in any analysis you run on Portfolio Visualizer, but like with MOAT, that doesn’t mean you should discount them entirely.

Lastly, consider the five-year returns for the 19 large-cap growth ETFs from 2014 to 2018. Again, there’s little new to note, as QQQ, IWY, and SPGP look great. This could be the one area where performance-chasing works.

Even so, I’m cautious not to discount any ETF entirely, as even recent poor performers like the Invesco S&P 500 Pure Growth ETF (RPG) experienced success before. Previously a Guggenheim fund, RPG ranked #2/11 on five-year returns for the year ending December 2011 and was the #1 ranked large-cap growth ETF for the five-year periods ending December 2012, 2013, and 2014.

Investment Recommendation

This analysis demonstrated that one-year, three-year, and five-year historical returns are not good predictors of future results. The only exception is in the large-cap growth category, where ETFs like QQQ and IWY routinely perform well. It remains to be seen if this success can continue, but it’s worth noting that several of the consistent-performing large-cap ETFs are high-quality plain-vanilla funds. Simple strategies, often with low expense ratios, might be the best solution for many investors.

I wrote this article because many readers and analysts rely primarily on past performance when deciding if an ETF is worth buying. The evidence presented today strongly indicates this approach is flawed. Solid long-term returns (i.e., ten years) are a bonus, but I will continue to focus my analysis on the fundamental strengths and weaknesses of an ETF’s underlying portfolio, which, fortunately, can be determined without performance charts. I hope you found this information helpful, and if you have any questions, please feel free to comment below. Thank you for reading.