Options income research
How I sell cash-secured puts and covered calls on stocks I want to own, and what two published studies say about the choices behind it.
- Status
- Research note
- Covers
- Cash-secured puts, covered calls
- Sources
- Goldman Sachs, 2013; Israelov and Nze Ndong, 2023
This page describes how I trade and what the research says about it. It shows no positions or results from my account. Every study figure below comes from the two papers; the payoff tool uses illustrative inputs you can change.
My approach
Which stocks
Only companies I would be glad to own. A put can turn into a purchase, so the stock has to pass the same test as an ordinary buy.
Put strike
A price at which I would happily buy the shares. If the put is exercised, I own a stock I wanted at a price I chose, less the premium already collected.
Call strike
Far enough above the share price that the premium is 0.1-0.5% of it. The aim is to keep the shares, so I accept a small premium for a low chance of having them called away.
Expiry
About one month.
What the trade pays
The two positions above, at expiry, against simply owning the shares. Everything is a percentage of today's share price, so no figure here depends on the size of an account.
The art of put selling
Goldman Sachs Options Research, April 2013. One-month puts on S&P 500 stocks from January 2003 to January 2013, fully backed by cash.
Selling puts earned close to the stock market's return with a third less volatility: 7.1% a year with 12% volatility, against 7.3% and 18% for the S&P 500 with dividends reinvested.
Where the strike sits
One-month puts, 2003 to 2013. Horizontal axis: how far below the share price the strike sits.
- Put selling
- S&P 500 total return
Return a year
Volatility
Share of puts exercised
Show the figures
| Strike below the share price | Return a year | Volatility | Sharpe ratio | Share of puts exercised |
|---|---|---|---|---|
| At the price | 7.1% | 12% | 0.65 | 45% |
| 2% | 6.6% | 11% | 0.66 | 35% |
| 5% | 5.9% | 9% | 0.68 | 23% |
| 10% | 5.5% | 7% | 0.80 | 12% |
| 15% | 5.0% | 6% | 0.85 | 10% |
| S&P 500 total return | 7.3% | 18% | 0.49 |
Returns include bid-ask spreads but not commissions. Source: Goldman Sachs, Exhibit 4.
Other results from the study
S&P 500 stocks, 2003 to 2013| Puts sold | Return a year | Sharpe ratio |
|---|---|---|
| On the fifth of stocks with the highest free cash flow yield | 9.8% | 0.85 |
| On the fifth of stocks with the lowest free cash flow yield | 4.3% | 0.36 |
| One month before expiry, at 50 delta | 7.1% | |
| Twelve months before expiry, at 50 delta | 6.3% |
In 2008, put selling beat the S&P 500 by 14 percentage points, because the premium cushioned the fall.
What it means for my approach
My put strikes sit below the share price, at a level I would pay, which is the right-hand half of the charts above. The data shows the trade-off: less income than selling at the price, but steadier results and far fewer assignments. Goldman's best screen used free cash flow yield as a margin of safety. My filter, companies I would own anyway, aims at the same thing but is less systematic.
A devil's bargain
Roni Israelov and David Nze Ndong, NDVR, October 2023. Covered calls on the S&P 500 index from January 1999 to June 2023.
Choosing a lower call strike raises the income but lowers the total return. A lower strike gives away more of the shares' gains, and the extra premium does not make up for it.
How much of the market's moves a covered call keeps
S&P 500 covered calls, 1999 to 2023. Horizontal axis: option income a year.
- Share of rises kept
- Share of falls kept
Show the figures
| Option income a year | Share of rises kept | Share of falls kept |
|---|---|---|
| None, shares only | 100% | 100% |
| 1% | 99% | 100% |
| 2% | 97% | 99% |
| 4% | 95% | 98% |
| 6% | 92% | 96% |
| 12% | 85% | 92% |
| 18% | 79% | 87% |
Upside and downside beta to the S&P 500. My premiums of 0.1-0.5% a month equal about 1-6% a year. Source: Israelov and Nze Ndong, Exhibit 11.
What the calls added to returns
A year, compared with holding the index| Option income a year | 1999 to 2023 | 2011 to 2023 |
|---|---|---|
| 1.2% and 2.4% | About 0.2% | A loss |
| 6% | (0.6%) | (3.1%) |
| 12% | (1.1%) | (4.7%) |
What it means for my approach
My 0.1-0.5% a month is about 1-6% a year, the low end of the range studied, where selling calls cost the least and at the very bottom slightly helped. A small premium keeps most of the upside, which fits the aim of holding on to the shares. The study used index options and I sell calls on single stocks, so its numbers are a guide rather than a forecast for my trades.
Why selling options can pay
Both studies trace the returns to two sources. Holding shares, or promising to buy them, earns the equity risk premium. Selling options earns the volatility risk premium: options are usually priced for bigger moves than actually happen. From 2003 to 2012, one-month S&P 500 options implied 20.1% volatility on average, while the index moved at 17.1%.
The two trades are close relatives. A put and a covered call at the same strike have almost the same payoff. Out of the money they differ: a put below the price cushions a fall better, while a call above the price keeps more of a rise.
Limitations
- Both studies use US large caps, and their data ends in 2013 and 2023.
- Goldman's returns include bid-ask spreads but not commissions, and the covered call study leaves trading costs out.
- Neither study tests my exact rules. Goldman set strikes by distance, delta or premium rather than a buy price, and the covered call study used index options rather than single stocks.
- Goldman Sachs does business with the companies it researches and NDVR manages money, so both authors have interests of their own.
Sources
- Marshall, J., K. Gregory and K. Fogertey. "The Art of Put Selling: A 10 year study." Goldman Sachs Options Research, 4 April 2013. Client research, not publicly available.
- Israelov, R. and D. Nze Ndong. "A 'Devil's Bargain': When Generating Income Undermines Investment Returns." NDVR, 26 October 2023. SSRN 4580048.