Passive income through option writing: Part 1

On the path to early retirement (and most likely in early retirement as well), the ERN family will be writing options to generate passive income (in addition to equity and real estate investments, of course). This may be something that people either haven’t heard before or even if they did, they might be turned off by the involvement of derivatives. After we got over our initial aversion against trading exotic instruments like options we found that it’s actually a reliable and profitable strategy to generate passive income. We mentioned this strategy already in a previous post on trading derivatives on the path to FIRE and thought that others might find this interesting too.

Today, in Part 1, we will do a quick intro to cover mostly the conceptual aspects of this strategy. Part 2 will go into how we actually implement our strategy. As a warm up, though, let’s start with a …

Pop Quiz:

Since 2000, the SPY ETF (S&P500 index fund from iShares) returned about 101% (Dec 1999 to August 2016, dividends reinvested), or about 4.3% p.a. What would the return have been if we had participated only when the market went up, i.e., if we had avoided every single down month and received a 0% return during that time?

A: 386% total, 10.0% annualized

B: 1,039% total, 15.7% annualized

C: 2,497% total, 21.6% annualized

D: 3,891% total, 24.8% annualized

The correct answer: D. Participating in all the monthly S&P500 gains but avoiding the negative returns would have generated a whopping 24.8% annualized return. $100 would have turned into close to $4,000 (see chart below); move over Warren Buffett! Whether it’s the folks in the FIRE community who are worried about the sequence of return risk, pension funds who have to fund their planned expenses or any other investor: everybody dreads the downside because it has robbed us of 20% of annualized returns recently!

 

sp500-cumul-returns
S&P500-TR (Total Return) comparison: With and without negative return months

How do we avoid the downside in practice? If I knew how to do this consistently and reliably, I wouldn’t be working for a paycheck anymore. And I would keep that secret to myself and not blurt it out here! Well, there is one way to prevent the downside: buy a put option with a strike price close to today’s index value. Finance lingo geeks would call that an “at the money option.” But that’s going to cost you! You have to pay a premium for that put option. If we look at the diagram below, these are return profiles as a function of the benchmark return:

  • The black dashed line is the 45-degree line (i.e., the benchmark itself).
  • The blue line is the profile that will get you the 24.8% average return mentioned above: you fully participate on the upside but avoid the downside.
  • The orange line is what you’d get when you buy a put option. You get exactly the blue return line, but you shift it down by the option premium. In this example, a little more than 1% of the underlying. We already see one potential problem with this method: you may not lose money on the equity index, but paying the option premium will pull your returns below zero. There is no free lunch: We will never be able to avoid all negative return months!
  • The yellow line is the payoff to the person who sold us the Put option.

putwriting-payoff-diagram1

Why doesn’t everybody buy put options as protection?

Ok, great! Let’s buy put options, hedge out the downside, lose at most the option premium every month but fully participate when the market goes up! If we could get that insurance for, say, 1% per month we still stand to gain 24.8%-12.0%=12.8% on average. Not bad!

Unfortunately, in reality, the insurance is quite a bit more costly. Let’s look at a concrete example. While writing this, I took the following market snapshot on September 21, at around the market close:

  • S&P500 future quote 2,155.75 (e-mini contract with a multiplier of 50)
  • A put option on this futures contract with a strike price 2,155 and expiration October 21 cost $28.75. Note that both the future and the option on the future have a multiplier of 50. So it would have cost us $1,437.50 (=$28.75*50) to ensure that our $107,787.50 (=$2,155.75*50) worth of equity holdings never drop below $107,750 (=$2,155.00*50) .
  • That’s a cost of about 1.33% to hedge the downside for the 30 days. Scaled up to an entire year that’s about 16.2%.

Bummer: That’s pretty darn close to the 20% excess return that the zero percent floor provides. It’s expensive to hedge the downside! And it gets even worse. That’s because right around the time when everybody is looking to buy insurance (2008, 2009, August 2015, January 2016, Brexit in June 2016), Put options become even more expensive. Much more expensive! If you take the average hedging costs over not just the relatively tranquil months like right now, but during all months since 2000 you will wipe out your entire expected equity return. The cost to insure the downside during August 2015 and January 2016 was around 30-40% annualized (!) and it would have been north of 80% annualized (~7.5-8.0% monthly !!!) in late 2008 during the Global Financial Crisis.

Geez, the people who sell put options are getting away with murder. Well, maybe not murder, but definitely highway robbery or racketeering! Who are the put option sellers, anyway? To enter that line of business you probably have bribe some politicians, right?

No! Meet Mr. and Mrs. ERN, put option sellers. It turns out that our brokerage account accommodates both option buying and selling. We will talk about the exact implementation next week, but in our experience, there was no major obstacle in setting up our account to sell options.

Introducing: The least appealing return profile

Of course, some people would argue that selling put options exposes you to the least appealing return profile, see yellow line in the chart above. There is an (essentially) unlimited downside, but only a limited upside. That’s the exact opposite of what everybody desires. But if the average premium is high enough to compensate for the occasional steep losses, who cares? This strategy can still be a winner!

Why the Short Put return profile is actually very attractive to us

I don’t even believe that the put option return profile is that unattractive. As a function of the benchmark return, let’s plot both the put writing strategy return (same yellow line as above) as well as the incremental return vis-a-vis the benchmark, see the diagram below. We can split the diagram into four distinct regions:

PutWriting Payoff Diagram2.png

  1. If the benchmark goes down by more than the option premium we will lose money as well, but the loss will be less than the benchmark loss. The put premium cushions the loss. Not very pleasant but better than losing as much as the benchmark.
  2. If the benchmark goes down by less than the option premium we do the happy dance! We make money while the benchmark lost. We actually hedged the risk in our other equity investments. Woo-hoo!
  3. If the benchmark is up, but by less than the option premium, we beat the benchmark. Yay! We are definitely happy campers!
  4. If the benchmark goes up by more than the put option premium we make money, though less than the benchmark. Is that a problem? We have tons of retirement account money invested in stocks, so it’s not that we’re missing out on the gains, we just don’t fully participate in the rally! We are still happy campers when that happens!

It turns out that in each of those regions I would be sitting pretty happy with my Short Put option. So the talk about how this is the worst possible return profile is vastly exaggerated.

Show me the money

How big is the premium in practice and how much return would we make if we consistently sold put options for the last 30 or so years? I consulted the CBOE who put together an interesting fact sheet and a research white paper about this strategy:

  • $1.00 invested on June 30, 1986, would have grown to $16.42 by January 29, 2016, if using the option writing strategy. Investing in the S&P500 your portfolio would have grown to only $14.83 (dividends reinvested). Despite selling the equity upside, writing puts would have returned more than the equity index. That’s how overpriced put options are. We say it again: it’s legalized highway robbery!
  • Exhibit 18, Page 9 in the white paper: Despite beating the S&P500’s compound return (10.16% vs. 9.85%, annualized, compounded), the put writing strategy had a lower volatility: 10.16% annualized (based on monthly returns), compared to the S&P’s 15.26%. So, the risk-adjusted returns of the put writing index was significantly higher: A Sharpe Ratio of 0.67 vs. the S&P’s 0.47. You get the same return for two-thirds the volatility. Sweet!

Markets are efficient. Then why should any of this option-writing business work?

Efficient markets are the reason why this works. When you sell puts you voluntarily subject yourself to the most undesirable and unattractive payoff profile possible: limited upside and unlimited downside potential. The exact opposite of what everybody wants. The efficient market compensates you for taking losses when they hurt the most.

Why is there not more supply of downside insurance? Nobody has the stomach anymore. Big institutional investors, like pension plans, are actually net buyers of downside protection. If you work for a pension fund or endowment you have a great aversion to downside risk because one big drawdown is all it takes for you to lose your job. Career risk for money managers creates demand for put options. I don’t have that career risk because I manage our own money. I’m not going to fire myself!

Banks face a lot of regulation and can’t afford to take on more risk and drawing the regulators’ scrutiny for a few percentage points of extra return per year. Hedge funds and private investors with a long-term focus and a stomach to sustain temporary short-term losses (and an understanding spouse!) seem to be the typical investors pursuing our strategy.

More information

Other bloggers have written about writing options, (both covered call writing and put shorting):

I also found the podcasts from the Options Industry Council very interesting, see the ones titled “Selling Puts.”

Also, check out the second part of this series. Passive income through option writing: Part 2: How we actually implement our option trading strategy.

Conclusion

Equity investors are compensated for taking on risk. You deserve a higher expected return from your equity portfolio than from a CD or money market account. But that compensation for taking on risk is exclusively for the downside. Investors who have no stomach for losses and only want capital gains are asking for a return profile that’s essentially a lottery or casino style payoff profile (small premium and large potential payoffs). It shouldn’t come as a big surprise that the expected return of such a strategy is also sub-par: just like in the casino.

Given that the average premium for this downside protection is actually quite rich, selling the insurance should be a good way to generate passive income. We have been doing this successfully since 2011. Simulations going back to 1986 also look very promising! After detailing our philosophy and confirming that this kind of strategy actually works in practice, let’s dive a little bit deeper into what exactly we are doing in next week’s post. Stay tuned!

We hope that we sparked your interest in options trading. Please leave your comments/questions/complaints below!

Advertisements

24 thoughts on “Passive income through option writing: Part 1

  1. Great explanation, ERN! Also looking forward to Part 2.

    Any thoughts on coupling your strategy with trading the VIX (probably VXX)? For instance, the option writing strategy you outline would, I think, tend to perform best (worst) during periods of peak market instability (stability), when VIX is highest (lowest). Possibilities for protection from the “massive downside scenario” with something like this? Just curious – admittedly haven’t given all this too much thought, but it popped to mind as I read your piece.

    Nicely done as always! Thanks!

    Liked by 1 person

    • Great question! Very thoughtful!!!
      It’s always the cost of protection that’s going to getcha. Going long the VXX is prohibitively expensive. The 5 year chart looks like the VXX went down 99%. Yup, -99% return in 5 years. You have massive contango in the VIX futures term structure and you constantly lose money when rolling to the next month once the current contract expiration gets closer. So much so that I would consider shorting the VIX (either the VXX ETF or the VIX futures). I prefer shorting the weekly put options, though. I get 52 independent bets per year. Only 12 bets with the monthly VIX futures.
      But I’ll think about it more and maybe write something in the future.
      Cheers!

      Liked by 1 person

  2. […] Financial engineering: Do a covered-call writing strategy, i.e., sell some or all of the upside potential of your stocks for extra yield. You still maintain the downside risk, but as we noted above, if the market really tanks, maybe it’s best to not jump into the housing market at that time anyway. Check Amber Tree Leaves for the basics on covered call writing. He also has a nice general intro to option trading. We also wrote a piece on option writing as a way to generate passive income. […]

    Like

  3. This is a really interesting post. I’ve read this post over the last three days over and over again trying to digest it and really understand. I’ll probably need to further review but it definitely looks like an interesting strategy in order to create passive income.

    Like

  4. Isn’t this just another version of the gambler’s fallacy? https://en.wikipedia.org/wiki/Gambler%27s_fallacy
    You’re expecting future volatility to be like past volatility when selling puts. What happens when there is some randome event. Also, isn’t this one of the main reasons LTCM lost so much money and had to be bailed out? Quoted from here – https://econ.duke.edu/uploads/assets/dje/2001/prabhu.pdf

    “LongTerm [Capital Management] took the position that historical volatility in the equity markets was an accurate indicator of future volatility. However, oftentimes their traders found that options traded at a
    price, which according to the Black-Scholes formula would imply a volatility well above the
    historical volatility of the underlying. The explanation that LTCM came up with for this
    divergence was that there was a greater demand for options than there was a supply. LongTerm
    felt that many investors, who were perhaps unsophisticated, were eager to obtain
    insurance for their portfolios, and therefore bid up the prices of equity options. According to
    the Black-Scholes formula, the volatility of the underlying asset and the price of an option on
    that asset are directly correlated. Therefore, increases in the prices of options due to the high
    demand for these securities were implicitly increasing the volatility implied by the options
    prices. Long-Term therefore sold options, which meant that they were implicitly selling
    volatility, a commodity that they believed had become overpriced.”

    “Value-at-Risk models used by LTCM rely on historical data to project
    information about future price movements. These models project the probability of various
    losses based on the prior history of similar events. Unfortunately, the past is not a perfect
    indicator of the future. On October 18, 1987, for example, two-month S&P futures contracts
    fell by 29%. Under a lognormal hypothesis, with annualized volatility of 20%
    (approximately the historical volatility on this security), this would have been a –27 standard
    deviation event. In other words, the probability of such an event occurring would have been
    10 raised to the -160. This is such a remote probability that it would be virtually impossible for it to happen. Similarly, on October 13, 1989 the S&P 500 fell about 6%, which under the above
    assumptions would be a five standard deviation event. A five standard deviation event would
    only be expected to occur once every 14,756 years. There are many other examples of
    abnormal market events happening with greater frequency than these models would lead one
    to expect. It would appear then, that lognormal models for expected returns do not fully
    account for these large losses, and that prior estimates of volatility may not be able to
    accurately predict future price movements. This reliance on a risk model that tends to
    underestimate the probability of large downward movements in securities prices may have
    led Long-Term Capital to be overconfident in its hedging strategies.”

    Liked by 1 person

    • Good point. Put options might be pricing in an event we have never seen in 100+ years. Hence the excess premium. I can’t prove that they are not and you can’t prove that they are. A similar argument, by the way, is sometimes used to justify the (Mehra and Prescott) equity premium puzzle: a catastrophic event we haven’t observed yet. I personally believe that for the put option premium there are fundamental reasons for the premium (myopic investors who have to hedge no matter what the price may be).

      For what it’s worth, some responses to the issues you mentioned:
      A 3x leverage out of the money put would have fared not too badly in 1998 according to my simulations. Whatever sank LTCM, a short leveraged put wasn’t the sole reason. At least not at 3x leverage.
      LTCM went under for a number of different reasons. The bet that had the biggest leverage was the on-the-run vs. off-the-run Treasuries. I think they levered that up by 30 to 1. I heard 100 to one but forgot the exact reference.
      Another reason for the LTCM failure was the illiquidity of the convergence trades where they couldn’t get out of their big positions without impacting the price. With my 16-20 short puts I will never get too big for the market.
      Another reason is that eventually LTCM got desperate to borrow money from some of the Wall Street players. Bad idea: they didn’t get the funding and in the end the same banks bet against LTCM, which made the mess even worse. Another problem I don’t plan to have. 🙂

      As I tried to convey before: 3x leverage does not mean that a 29% drop gets close to wiping out the portfolio. Leading up to the big events in 1987, 1998, 2008 implied volatility was already very elevated when you sold those of the money puts. The 3x leverage doesn’t begin until you hit the strike price, significantly below the current price.
      On top of that, our marginal tax is over 35%, so the after tax leverage is actually less than 2.0.

      Like

  5. Thanks for the followup reply. I agree with your counter argument and think people need to be absolutely aware that this isn’t a “free lunch” investment where they can lever up 10 or 20x on a short put. People do outlandish things thinking there’s “no way” for this investment to lose. You’ve done a good job outlining one instance where your investment went bad. It would be nice for you to post a writeup of how the 3x short put would have done in ’87, ’98, or ’08 (maybe even versus a 1x and 5x).

    Great work on the blog so far! The amount of time that it must take you to make those nice figures alone makes me wonder if you’ve got a staff working for you.

    Liked by 1 person

    • Oops, hit the reply button too early, while still typing.

      Thanks for the compliment! Yes, I got people working for me at the office, but it would be unethical if I have them work on the blog, though, 🙂

      Regarding the drawdowns in the years you are interested in, the caveat here is that I have clean option pricing data only since 2003. Before then I would have to make assumptions about the shape of the vol smile curve (how much does implied vol go up whe you go more and more out of the money with your Put options). Before 1990 I don’t even have the VIX, so I would have to deduce that from daily return vol.
      with all the caveats: a 3x Put option as described in my other post would have done as follows:
      2008: you would have lost money in the weeks ending 10/10/2008 and 5/23/2008. In both cases the 3x option strategy would have lost less than the index. -9% in October, vs. -18% in the index. -0.1% in May vs. -3.5% in the index.
      That’s because the strike price was so far out of the money.
      Intriguingly, in the week of the Lehman failure you would have lost no money on a Friday to Friday basis. You would have lost on that Monday 9/15/2008, just because the time value went up. But all puts would have expired worthless and you would have made the max premium.

      Other years to follow soon…

      Like

    • 1998:
      during the volatile weeks in 1998 from late July to early October you would have lost only during that first week ending on 7/24/1998. (This with the caveat that I have no option data, but rather have to deduce how far out of the money I would have sold the options, considering the VIX level at the Friday close. I’m not even sure they had weekly options back then)
      Options were far enough out of the money to avoid a deep drawdown even that week: -1.5% even at 3x leverage. In the weeks following you would have made the maximum premium. With the VIX around 40 you should have been able to sell the puts 10% out of the money, according to my “vol smile model”
      So, 1998 was a year when this strategy worked really, really beautifully.

      Like

    • 1987, October
      Quite an extraordinary month, because the options strategy lost money three weeks in a row. Normally, a large loss one week sends the VIX up high enough that the strike price for next week is so far out of the money, it would be unlikely for the SPX to fall below 2 or even three times in a row.
      In the weeks ending on the 9th, 16th and 23rd of October 1987, the index went down 5%, 9%, 12% and the option strategy 2%, 10%, 7%, respectively. It took over a year to recover from that drawdown.
      What’s even worse: the intra-week performance was pretty horrendous. You were lucky that on 10/16 the SPX dropped by 5%, so you should have been able to sell the put options about 10% out of the money. But when the market dropped by 20% on 10/19 you still lost 30% with your 3x options. The recovery later that week brought your loss back to single digit percent, but the risk was that you lost you nerve during the rout on Monday and missed the recovery on Tuesday and Wednesday (10/20 and 10/21).

      So, yeah, I’m the first to admit that October 1987 doesn’t look good. The one piece of good news is that these kinds of events normally don’t occur out of nowhere. There is usually some volatility and fear building up beforehand so you can sell at strikes far, far out of the money.

      Like

  6. I’ve been thinking of doing something similar since growing my sophistication enough to put in limit orders well below current market price so as not to miss a flash crash because I was too busy working or playing.

    Selling puts would be essentially the same thing, wouldn’t it, except that I’d get paid to provide that insurance for others?

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s