{"id":1371,"date":"2024-09-04T20:21:16","date_gmt":"2024-09-04T20:21:16","guid":{"rendered":"https:\/\/www.allendowney.com\/blog\/?p=1371"},"modified":"2024-09-04T20:21:16","modified_gmt":"2024-09-04T20:21:16","slug":"rip-off-etf","status":"publish","type":"post","link":"https:\/\/www.allendowney.com\/blog\/2024\/09\/04\/rip-off-etf\/","title":{"rendered":"Rip-off ETF?"},"content":{"rendered":"\n

An article in a recent issue of The Economist<\/em> suggests, right in the title, \u201cInvestors should avoid a new generation of rip-off ETFs<\/a>\u201d. An ETF is an exchange-traded fund, which holds a collection of assets and trades on an exchange like a single stock. For example, the SPDR S&P 500 ETF Trust (SPY) tracks the S&P 500 index, but unlike traditional index funds, you can buy or sell shares in minutes.<\/p>\n\n\n\n

There\u2019s nothing obviously wrong with that \u2013 but as an example of a \u201crip-off ETF\u201d, the article describes \u201cdefined-outcome funds\u201d or buffer ETFs, which \u201coffer investors an enviable-sounding opportunity: hold stocks, with protection against falling prices. All they must do is forgo annual returns above a certain level, often 10% or so.\u201d<\/p>\n\n\n\n

That might sound good, but the article explains, \u201cOver the long term, they are a terrible deal for investors. Much of the compounding effect of stock ownership comes from rallies.\u201d<\/p>\n\n\n\n

To demonstrate, they use the value of the S&P index since 1980: \u201cAn investor with returns capped at 10% and protected from losses would have made a real return of 403% over the period, a fraction of the 3,155% return offered by just buying and holding the S&P 500.\u201d<\/p>\n\n\n\n

So that sounds bad, but returns from 1980 to the present have been historically unusual. To get a sense of whether buffer ETFs are more generally a bad deal, let\u2019s get a bigger picture.<\/p>\n\n\n\n

Click here to run this notebook on Colab<\/a> <\/p>\n\n\n\n

The Dow Jones<\/h2>\n\n\n\n

The MeasuringWorth Foundation<\/a> has compiled the value of the Dow Jones Industrial Average at the end of each day from February 16, 1885 to the present, with adjustments at several points to make the values comparable. The series I collected starts on February 16, 1885 and ends on August 30, 2024. The following cells download and read the data.<\/p>\n\n\n\n

DATA_PATH = \"https:\/\/github.com\/AllenDowney\/ThinkStats\/raw\/v3\/data\/\"\nfilename = \"DJA.csv\"\ndownload(DATA_PATH + filename)\n<\/pre>\n\n\n\n
djia = pd.read_csv(filename, skiprows=4, parse_dates=[0], index_col=0)\ndjia.head()\n<\/pre>\n\n\n\n
<\/th>DJIA<\/th><\/tr>
Date<\/th><\/th><\/tr><\/thead>
1885-02-16<\/th>30.9226<\/td><\/tr>
1885-02-17<\/th>31.3365<\/td><\/tr>
1885-02-18<\/th>31.4744<\/td><\/tr>
1885-02-19<\/th>31.6765<\/td><\/tr>
1885-02-20<\/th>31.4252<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
To compute annual returns, we\u2019ll start by selecting the closing price on the last trading day of each year (dropping 2024 because we don\u2019t have a complete year).<\/p>\n\n\n\n
annual = djia.groupby(djia.index.year).last().drop(2024)\nannual\n<\/pre>\n\n\n\n
<\/th>DJIA<\/th><\/tr>
Date<\/th><\/th><\/tr><\/thead>
1885<\/th>39.4859<\/td><\/tr>
1886<\/th>41.2391<\/td><\/tr>
1887<\/th>37.7693<\/td><\/tr>
1888<\/th>39.5866<\/td><\/tr>
1889<\/th>42.0394<\/td><\/tr>
…<\/th>…<\/td><\/tr>
2019<\/th>28538.4400<\/td><\/tr>
2020<\/th>30606.4800<\/td><\/tr>
2021<\/th>36338.3000<\/td><\/tr>
2022<\/th>33147.2500<\/td><\/tr>
2023<\/th>37689.5400<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
139 rows \u00d7 1 columns<\/p>\n\n\n\n
Next we\u2019ll compute the annual price return, which is the ratio of successive year-end closing prices.<\/p>\n\n\n\n
annual['Ratio'] = annual['DJIA'] \/ annual['DJIA'].shift(1)\nannual\n<\/pre>\n\n\n\n
<\/th>DJIA<\/th>Ratio<\/th><\/tr>
Date<\/th><\/th><\/th><\/tr><\/thead>
1885<\/th>39.4859<\/td>NaN<\/td><\/tr>
1886<\/th>41.2391<\/td>1.044401<\/td><\/tr>
1887<\/th>37.7693<\/td>0.915861<\/td><\/tr>
1888<\/th>39.5866<\/td>1.048116<\/td><\/tr>
1889<\/th>42.0394<\/td>1.061960<\/td><\/tr>
…<\/th>…<\/td>…<\/td><\/tr>
2019<\/th>28538.4400<\/td>1.223384<\/td><\/tr>
2020<\/th>30606.4800<\/td>1.072465<\/td><\/tr>
2021<\/th>36338.3000<\/td>1.187275<\/td><\/tr>
2022<\/th>33147.2500<\/td>0.912185<\/td><\/tr>
2023<\/th>37689.5400<\/td>1.137034<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
139 rows \u00d7 2 columns<\/p>\n\n\n\n
And the relative return as a percentage.<\/p>\n\n\n\n
annual['Return'] = (annual['Ratio'] - 1) * 100\n<\/pre>\n\n\n\n
Looking at the years with the biggest losses and gains, we can see that most of the extremes were before the 1960s \u2013 with the exception of the 2008 financial crisis.<\/p>\n\n\n\n
annual.dropna().sort_values(by='Return')\n<\/pre>\n\n\n\n
<\/th>DJIA<\/th>Ratio<\/th>Return<\/th><\/tr>
Date<\/th><\/th><\/th><\/th><\/tr><\/thead>
1931<\/th>77.9000<\/td>0.473326<\/td>-52.667396<\/td><\/tr>
1907<\/th>43.0382<\/td>0.622683<\/td>-37.731743<\/td><\/tr>
2008<\/th>8776.3900<\/td>0.661629<\/td>-33.837097<\/td><\/tr>
1930<\/th>164.5800<\/td>0.662347<\/td>-33.765293<\/td><\/tr>
1920<\/th>71.9500<\/td>0.670988<\/td>-32.901240<\/td><\/tr>
…<\/th>…<\/td>…<\/td>…<\/td><\/tr>
1954<\/th>404.3900<\/td>1.439623<\/td>43.962264<\/td><\/tr>
1908<\/th>63.1104<\/td>1.466381<\/td>46.638103<\/td><\/tr>
1928<\/th>300.0000<\/td>1.482213<\/td>48.221344<\/td><\/tr>
1933<\/th>99.9000<\/td>1.666945<\/td>66.694477<\/td><\/tr>
1915<\/th>99.1500<\/td>1.816599<\/td>81.659949<\/td><\/tr><\/tbody><\/table><\/figure>\n\n\n\n
138 rows \u00d7 3 columns<\/p>\n\n\n\n
Here\u2019s what the distribution of annual returns looks like.<\/p>\n\n\n\n
from empiricaldist import Cdf\n\ncdf_return = Cdf.from_seq(annual['Return'])\ncdf_return.plot()\n\ndecorate(xlabel='Annual return (percent)', ylabel='CDF')\n<\/pre>\n\n\n\n
\"\"<\/figure>\n\n\n\n
Immediately we see why capping returns at 10% might be a bad idea \u2013 this cap is exceeded almost 45% of the time, and sometimes by a lot!<\/p>\n\n\n\n
1 - cdf_return(10)\n<\/pre>\n\n\n\n
0.4492753623188406\n<\/pre>\n\n\n\n
Long-Term Returns<\/h2>\n\n\n\n
We\u2019ll use the following function to compute long-term returns. It takes a start date and a duration, and computes two ratios:<\/p>\n\n\n\n