Does It Qualifies as Mechanical Investment?

Investing Strategies Mechanical Investing
1

And, is it any good?

This is a Covered Call (CC) strategy, so if you hate them skips to the next post. Also, I won’t review here what a CC strategy is as there myriads of videos and write ups on the subject.

In general, CC strategies do not sit well with seasoned investors as they cap the upside gains and provide limited protection to the down side in the form of option premium. Why bother with a 1-2% monthly return when the market is falling like a brick? That’s true for the most part. But given that option premium are linked to implied volatility (iVol), and markets are experiencing turmoil, the VIX which is one way of reflecting iVol, is at a mid level of ~25-29. Thus, option premiums are higher than usual.

For example, Wells Fargo (WFC) is paying 1.80$ for At-The-Money (ATM) Calls expiring 4 weeks down the road. With the share price at 40$, that’s 54% (1.80x 12 /40) potential annual return, before changes in share price. For comparison, the famous QYLD ETF, that pays annual “dividends” of ~12% from ATM option premium looks like it’s ripping off investors as a similar approach using QQQ can yield +4% monthly return, without paying the 0.6% annual fees.

Can we do better? In April 2022 CME launched daily options for the E-Mini futures. So now there are 5x/week expiring option chains as opposed to the previous 3x/wk calendar. In principle, at 3:50 pm ET every day one can write ATM calls against E-Mini futures (ES), valued @3,800 while getting ~22/day. Here we are talking a high 145% max potential annual return (22 x5days x50 weeks /3800) based on option premium, again before including share price variation. These numbers are just a nominal reference as futures are purchased on margin. For example, in a portfolio margin account a CC position in ES requires 20M$. The leverage here is significant as one Mini-ES futures has a notional value of 50x the index, right now @3,800x50= $190,000. So, the theoretical max returns on margin are much higher: (22x50x250)/20,000 = 1,375%. Careful with theses numbers as CC are net long and the leverage plays against you just as much. More on risk latter.

Results
The Chicago Board Options Exchange (CBOE) came up with a monthly CC index based on SPX with data from 2002. It’s called BMX, where one buy the SPX index and then writes ATM call options.
https://www.cboe.com/us/indices/dashboard/BXM/
The results are quite interesting. Starting in 2002, it took SPX until 2017 to surpass BXM, assuming one reinvest the Option premium. Not bad for a strategy viewed as having limited upside potential.

And, they use monthly options which provides significantly less premium over a 12 month period. How much less versus the new dailies? Over 4x at equivalent VIX.

How soon can we recuperate from a market decline? Another way of looking at risk. Let’s use an example pull out of the hat. In this bear market, the SPX bottoms at 2,000, overshooting the march 2019 low of 2174. From the current 3,800-2,000 = 1,800 pts x 50$/point there is -$90,000 loss for each Mini-ES contract. At the current daily premium of 22, it would require 82 days (1,800/22) to cover that amount. The option multiplier for ES options is 50, just like the outright contract. By the way, the current 22 in option premium would certainly move much higher as the VIX creeps up. Again, no bad for strategy that’s usually perceived as having little downside protection.

The risk increases as the speed of decline accelerates since there are not enough days to gain that protective premium. But, quick downturns like the one in March 2020 can be overcome with Tail Hedging for as low as 1-1.5%/year as the explosive gains in iVol fuel Out-The-Money protective puts. But that’s another post.

As I was writing this note the VIX increased by 2.4pts to 29.8 and the ATM calls option expiring tomorrow went from 22 to 30.

To close. Is it a MI strategy? I guess so as it’s pretty mechanical. Everyday before the market ends, one close the current day positions and open a new one for next day. Rinse and repeat everyday. A small twist is to use only naked calls, instead of the usual CC combination, when the market is below the 200 day moving average and therefore falling. This changes the profile of the position into a net short. And move back to regular CC when the market is above the 200 SMA.

Any thoughts? What am I missing?

7 Likes
2

Any thoughts? What am I missing?

Regarding BMX, I suspect that you’re missing start/end point effects. Perhaps look at the difference of rolling one year returns to see if there’s a sustained advantage.

CBOE has a ton of option strategies for which they publish backtests. Some are fairly fancy like collars, condors etc. None of the ones that I’ve analyzed are particularly attractive.

Given that their strategies are based on heavily traded indices, perhaps this isn’t surprising - an advantage would quickly be arbitraged out.

3

Perhaps surprisingly the covered call indexes outperformed the underlying benchmark for a whole lot of years,
but as noted, not always, and in particular not in the last few years.

The thing to remember is that covered calls have a very particular set of situations when they do well.
A covered call position will outperform the underlying long position if the underlying security falls in price a lot, falls a little, stays flat, or rises a little.
The covered call will underperform the stock if the underlying rises a lot. (though the CC will still be profitable)

This is true for each individual time interval of an option.
Repeated over time, the chances get better that you’ll get an advantage:
The observation that nothing rises a lot in every time interval is the reason that covered calls often perform quite well over time.

I have almost never done covered calls, but I have done a whole lot of cash-backed put writing, about 100,000 contracts.
The two strategies are not NEARLY as equivalent as the textbooks suggest, but alike in this particular way:
There is a price for everything.
With a tiny premium it doesn’t make sense. With a huge premium it does. Premiums vary a lot over time.
Pick your moments and you’ll do well.

So, put together, the rules are:
If premiums are currently high (e.g., VIX decently high), and the underlying is not currently soaring in price or expected to soar in price imminently,
a covered call portfolio is likely to outperform a long portfolio of the same underlying securities.

If either or both of those tests is wrong, the equation shifts. If both are wrong, don’t do it.
So, for example summer last year, the VIX was in the basement and the markets were soaring in a smooth seemingly risk free way.
The worst time to be holding covered call positions.

So my suggestion:
You can probably get a huge advantage with a basic test for whether you’re in an ongoing strong bull market,
and checking the implied return from the option before entering the position.
Simply use another strategy when those two omens are bad, and I think you’ll likely find that
a covered call strategy will nicely outperform a long strategy the rest of the time.

It’s probably one of those situations that timing signals will help a lot.
They don’t have to be right all the time to add a lot of value, as long as the bullish periods are on average better than the bearish ones.
And the premium for the option is an easier test: you know the maximum possible rate of return before you enter the position.

Jim

9 Likes
4

I agree with Jim that it sounds plausible that you could time when to use one of these strategies.
For example, invest in SPY in bull markets and sell covered calls in bear markets (VIX is typically high in bears).

But at least in my hands such a plausible sounding strategy wasn’t borne out in practice at least using indices as the underlying.
I think this is because any potential advantages of a simple options overlay on indicies tends to get arb’d away.
I suspect a significant part of Jim’s success in put-selling (and Cohen’s) was due to the underlyings not being heavily traded indices.

No Timing
Here’re the stats for SPY and BMX from 1993-08-13 to present (after deleting one day that had a Yahoo data error):


                                BMX          SPY
Annualized Return            0.0801000    0.1117000
Annualized Std Dev           0.2297000    0.1904000
Annualized Sharpe (Rf=0%)    0.3487000    0.5866000
maxDrawdown                  0.6080976    0.5518943

CONCLUSION:
Based on just those stats, BMX isn’t looking attractive.
One should never draw general conclusions from a set of stats over one time period like the above.
Instead, I often look at the difference of rolling 1 year returns (also 3 and 5 years) in order to check consistency of over/under performance.
But I sort of recall doing a more intensive analysis in the past that didn’t change the conclusion.

Timing based on Simple Moving Average
Now let’s invest in BMX i.e. do covered calls, but only if SPY is below it’s SMA200 (bear) and otherwise invest in SPY (bull):


                             BMX_timed       SPY
Annualized Return            0.1084000      0.1117000
Annualized Std Dev           0.1883000      0.1904000
Annualized Sharpe (Rf=0%)    0.5758000      0.5866000
maxDrawdown                  0.6210535      0.5518943

CONCLUSION:
Based on just those stats, BMX_timed using SMA200 isn’t looking attractive.

I tried timing using VIX, i.e. above/below it’s third quartile value, and similarly saw no advantage.

But I’m just a dog with some time to spare this morning (actually, my dog is at the vet).

1 Like
5

But at least in my hands such a plausible sounding strategy wasn’t borne out in practice at least using indices as the underlying.
I think this is because any potential advantages of a simple options overlay on indicies tends to get arb’d away.
I suspect a significant part of Jim’s success in put-selling (and Cohen’s) was due to the underlyings not being heavily traded indices.

I presume you’re right about options on the index (or SPY, which amounts to the same thing)

It’s hard to say whether it’s a market inefficiency or not,
but mathematically any collection of stocks is going to be less volatile then any average stock in the collection.
Some of the noise cancels out, making any basket a smoother rid.
(This is arguably the only reason that investment funds exist: people pay a fee to have brokerage statements that are less scary.)
Presumably this means that the index options are cheaper.
Other than at panicky times that there is a premium for index puts due to fear, meaning prices are set more by supply/demand than by Black-Scholes.

I’ve never tested it, but this observation about aggregation would tend to lead to the conclusion that at least one of these surprising things is true:
(a) the prices are different, so there is free money to be made buying options on the index and selling options on the underlying individual stocks of the same basket;
(b) the prices aren’t different, so index options are forever too expensive; or
(c) the prices aren’t different, so individual stock options are forever too cheap.

Yes, most of my trades have been on individual stocks, specifically those I consider cheapish.
The best are firms that I think will be worth a lot more, but the market really hates, for a long time.

Jim

8 Likes
6

Thanks for the input.

The results presented by tedthedog using BMX vs SPY suggest that CC is not worth pursuing. But, let’s remember that BMX is based on monthly calls. Now there are daily options for the Mini-ES futures. And, at parity VIX, daily calls provide over 4x the premium received by the monthly call BMX index. That should skew the results in favor of CC vs long SPY to make it worthwhile.

Perhaps the right approach when the market is below the SMA200MA is not to go long with CCs, but rather short with naked Calls and use the option premium to protect against the upside. And, stop or reverse the approach, back to CCs, when the NHNL indicators fires a minor or major bottom.

At a conceptual level, as long as the implied volatility is higher than the historical one, writing options provide an edge vs buying them. I guess the analogy is that as an option writer one is selling insurance, and the long buyers are our customers. The house wins.

In general I can see CC being useful under the conditions outlined by Jim. Not worth it under runaway markets and low VIX. How low? I venture <20 or when the annualised option premium falls below 100% the stock/index price.

1 Like
7

ted’s analysis had a larger max drawdown with a timed BMX which seems impossible to me. Should probably double check the calcs.

8

But, let’s remember that BMX is based on monthly calls…

That is certainly an issue.
The shorter the time frame, the more price movements are random and a B-S model becomes a reasonable assumption.
The longer the time frame, the more hilarious the model becomes.

My own strategy (cash backed put writing, not covered calls) involved rules of thumb such as:

  • Pick underlying securities that I believe to be attractively valued that you wouldn’t mind owning.
    Think of the position not as a wager in the options market, but as just another way to be long the stock.
  • Lean towards the longest dated / lowest strike options that allow for a minimum 15%/year return from time value alone.
  • When stock prices rise, close and replace positions which no longer have a minimum acceptable maximum remaining annualized rate of return. Maybe 6%.
  • Don’t do it once. Pick a security, write them over and over for years if you can.
    Some short intervals will work out badly, but for any business not falling in value you’re very likely to do better than the stock over a longer time frame.
  • This lends itself to a bit of leverage: $100 in cash can easily back up $150 in short puts.
    Nobody exercises a put that they could sell for a bid that would realize any time value.
    One of the advantages of leaning towards longer time frames: more time value in your average position to start with, and staggered expiration dates.

Jim

9 Likes
9

Good one! The calculation was fine, the data wasn’t.

The OP wrote:
It’s called BMX, where one buy the SPX index and then writes ATM call options.
I grabbed Yahoo data for BMX.

The index is actually BXM (ticker ^BXM)
So one more time using the index data from Yahoo, the stats below are from 1993-11-13 to 2022-07-06


                             BXM_timed        SPY
Annualized Return            0.0941000      0.0965000
Annualized Std Dev           0.1644000      0.1904000
Annualized Sharpe (Rf=0%)    0.5723000      0.5068000
maxDrawdown                  0.4579059      0.5518945
1 Like
10

The index is actually BXM (ticker ^BXM)…

I didn’t know that Yahoo had the data for the index.
Is that, like most indexes, the price-only return, excluding the dividends on the portfolio?
The Eaton Vance ETV buy-write fund has outperformed that index by 3.74%/year since inception in 2005.
SPY has done better, but only by less than a percent–probably very endpoint dependent.
It was ahead from inception to end dates in the range 2009-2019, and did better in the last year.
The return profile is weirdly similar–you wonder why they bother. For the fees, of course.
Same standard deviation of and worst rolling years.

Also—

If I remember correctly, there are two related buy-write indexes.
One does at the money, the other does first out of the money, something like that.
One has done a lot worse than the other.
Anybody check that?

Also also—
There’s an interesting 2017 paper from AQR.
"The CBOE PutWrite Index has outperformed the BuyWrite Index
by approximately 1.1 percent per year between 1986 and 2015.
That is pretty impressive. But troubling. Yes – troubling –
because the theory of put-call parity tells us that such
outperformance should be almost impossible via a compelling
no-arbitrage restriction. This paper explains the mystery
of this outperformance, which has implications for portfolio
construction."
The reason for the difference is surprising and bizarre, though interesting only to numbers geeks.
It’s all about the details of the methodology the index is simulating, and nothing about the difference in puts versus calls.

Jim

4 Likes
11

I didn’t know that Yahoo had the data for the index.
It does.
I took a quick look at the CBOE site and what I saw there for a short period agreed with the Yahoo data.

Is that, like most indexes, the price-only return, excluding the dividends on the portfolio?
No, it apparently includes dividends:
From https://cdn.cboe.com/api/global/us_indices/governance/BXM_Me…

"Dividends paid on the component stocks underlying the S&P 500 Index and the dollar value
of option premium deemed received from the sold call options are functionally “re-invested” in the
covered S&P 500 Index portfolio."

1 Like
12

Jim, you get the ‘nerd of the month’ award!
From the paper referenced:

… the primary reason why the PutWrite Index has outperformed the BuyWrite Index is a construction difference during just four hours per month.
A quirky difference in their portfolio construction results in the PutWrite Index missing out on approximately four hours per month of S&P 500 Index return relative to the BuyWrite Index.

3 Likes
13

Jim:
regarding your put writing strategy, what do you do if stock price declines and the puts go in-the-money?
Do you take ownership of the stock and then sell calls?

14

Jim, How do you calculate time value, using margin or other?

My own strategy (cash backed put writing, not covered calls) involved rules of thumb such as:

  • Pick underlying securities that I believe to be attractively valued that you wouldn’t mind owning.
    Think of the position not as a wager in the options market, but as just another way to be long the stock.
  • Lean towards the longest dated / lowest strike options that allow for a minimum 15%/year return from time value alone.
  • When stock prices rise, close and replace positions which no longer have a minimum acceptable maximum remaining annualized rate of return. Maybe 6%.
  • Don’t do it once. Pick a security, write them over and over for years if you can.
    Some short intervals will work out badly, but for any business not falling in value you’re very likely to do better than the stock over a longer time frame.
  • This lends itself to a bit of leverage: $100 in cash can easily back up $150 in short puts.
    Nobody exercises a put that they could sell for a bid that would realize any time value.
    One of the advantages of leaning towards longer time frames: more time value in your average position to start with, and staggered expiration dates.

Jim

15

regarding your put writing strategy, what do you do if stock price declines and the puts go in-the-money?
Do you take ownership of the stock and then sell calls?

If it’s assigned, I sell the stock and immediately write a fresh put. That’s simple enough.

Until then I wait it out, generally.
When the put is well in the money the return profile is much like owning the stock, which I decided when I started was something I’d be OK with.

When it comes time to replace the options, due to assignment or just getting close to expiry,
the main decision is whether to stick with the same high strike which has little time value but guarantees I can’t be whipsawed,
or switch to a lower strike with more time value but with which I have to be more nimble if the stock bounces.
Since I watch the portfolio closely, I usually go for the latter. Sometimes in between, like slightly in the money.
The idea is this: I snag another big slug of time value, but if the price bounces and I make a big
profit on that time value quickly I replace it with a higher strike so I don’t let the stock “get
away” from me with ever dwindling return from my now far out of the money contracts.

But there are quite a few alternatives, so it depends how I feel about the stock.
If I think the stock has gone from cheap to really cheap, I might just write more puts.

Jim

5 Likes
16

Jim, How do you calculate time value, using margin or other?

Stock price = P
Strike price = S
The bid for the contract = B
So, I assume that I can sell for B, meaning if assigned my entry price is (S-B) = my max capital at risk. (that’s what I’d lose if the stock went to zero)
Assuming it is out of the money, the maximum profit is the same as the bid, B.
So the raw return is B/(S-B).
I then annualize that by multiplying by 365/(days to expiry).
I hope for more, but never do it for under 15%/year rate.
I don’t try to maximize this…I want the biggest margin of safety (long date, low strike) that gets me a decent number, say 15-22%/yr.
The nice thing is you can do all this math before you enter a position, and just say no if it isn’t interesting.
Using a limit, as one should always do, usually you do a little better than getting just the bid, but I don’t count on it.

I don’t assume any leverage when assessing a position.
Going into the position you need to have availability of (S-B) in cash in case of assignment.
(after you collect the premium, you need just S of course).

But it’s not too crazy to assume you’ll never have all contracts assigned the same day.
So though I ensure that I have enough for most contracts to be assigned the same day, you can get by with less than enough for ALL of them to be assigned.
(especially since worst case you’d use broker margin for a couple of hours till you sold the stock and replaced it with fresh puts)

Unless the maximum remaining rate of return is very low, I don’t close positions early.
I prefer to have them assigned, as that means I don’t have the horrible option contract bid/ask gap to pay to close the position.
Trading the stock is much easier and cheaper, so I prefer assignments.
Ideally assigned with just a few cents in the money, but then a nice bounce Monday morning before I sell it : )

I find that over time the return you get from repeated cash-backed put writing (without leverage)
will tend to be about halfway between the return on the stock in that time period, and around 10%/year rate.
If the stock is returning 10%, it’s a wash.
If the stock returns 16%, you might make 13%.
If the stock breaks even, you might make 5%.
If the stock loses 10%, you might break even.
A bit better if your picks are good, and/or if you are really careful about not doing it when the premiums on offer aren’t good.
Price matters a whole lot:
Picking up pennies in front of steamrollers is a bad idea.
But picking up gold coins in front of Tonka trucks is fine.

Jim

12 Likes
17

The idea is this: I snag another big slug of time value, but if the price bounces and I make a big
profit on that time value quickly I replace it with a higher strike so I don’t let the stock “get
away” from me with ever dwindling return from my now far out of the money contrac

Jim, I am confused about how to determine time value. Please explain

18

Jim, I am confused about how to determine time value. Please explain

If any option would not be immediately profitable if exercised by the holder (the person who bought it) it is out-of-the-money and the whole of its market value is time value.

For an out-of-the-money put option (low strike price, below the current stock price),
the time value is the entire market value of the option.
Since you’re trying to sell one of those, it’s the bid price that matters.

Since you’re trying to make a profit, what matters is the ratio of this time value to your capital at risk.
Better to sell a $2 option that ties up $5 than to sell a $10 option that ties up $100.

The time value versus intrinsic value is easier to explain with call options.
A call option is exactly like a rain cheque at the grocery store.
Imagine you had a rain cheque to buy a box of Cheerios for $2 any time in the next year, and the current price is $2.25.
You could use that coupon immediately and sell the box for $2.25, so it has an intrinsic value of $0.25 so it is an “in the money” option.
However, the market price of that rain cheque is probably more like $0.30.
That’s because somebody holding onto it doesn’t have to use it right away…maybe the price of Cheerios will be $5 next year?
They can decide to wait and make a killing, while risking very little in the mean time, so its value is higher than the immediately available profit.
So the ability to decide later, the optionality, is worth some money.
So in this case, the market value of the option is $0.30, the intrinsic value is $0.25 (difference between strike price $2 and price of the underlying asset $2.25), and a time value of $0.05.

Now, maybe you have an option to buy that same box for $2 but the current price of Cheerios is $1.90.
There is no intrinsic value to this option (in the options market sense of the term): no profit if it’s used immediately.
Maybe the market value of that option is worth $0.07 or something like that.
(again, maybe the price will be $3 in a year…you could lock in the $2 price now, so that optionality is worth something).
The market value of this option—and out-of-the-money-option—is entirely time value.
The intrinsic value is zero (no profit to be had if exercised immediately), and the time value and market value are both $0.07.

The value of any option is its intrinsic value plus its time value.
The time value of any option is highest when it has the most time to expiry, and when the strike price equals the stock price and the intrinsic value is zero.
The time value will evaporate with time, and will fade relatively quickly as the stock price moves away from the strike price.
In one direction the intrinsic value stays zero, in the other direction the intrinsic value rises dollar for dollar with the stock price.
Since the price is the sum of the two, the price of an option is the sum of a rough bell curve (peak time value in the middle when stock price at strike),
plus a half of a diagonal line (intrinsic value rises linearly from zero in one direction).
https://www.mathfinance.com/wp-content/uploads/2019/07/2019-…

Jim

11 Likes
19

I don’t understand why margin required for the trade is not used in the calculation, as your cost for the trade period, amount of your money tied up in the trade to calculate time value.

Stock price = P
Strike price = S
The bid for the contract = B
So, I assume that I can sell for B, meaning if assigned my entry price is (S-B) = my max capital at risk. (that’s what I’d lose if the stock went to zero)
Assuming it is out of the money, the maximum profit is the same as the bid, B.
So the raw return is B/(S-B).
I then annualize that by multiplying by 365/(days to expiry).
I hope for more, but never do it for under 15%/year rate.
I don’t try to maximize this…I want the biggest margin of safety (long date, low strike) that gets me a decent number, say 15-22%/yr.
The nice thing is you can do all this math before you enter a position, and just say no if it isn’t interesting.
Using a limit, as one should always do, usually you do a little better than getting just the bid, but I don’t count on it.

20

I don’t understand why margin required for the trade is not used in the calculation…

There is no margin required for cash-backed put writing, as such.
The cash pile provides all the security you need. You aren’t borrowing anything.

It changes a bit if you use a bit of leverage: more aggregate assignment value of your puts than cash on hand.
If you write puts with an aggregate notional value somewhat higher then your cash pile, you are tying up some of the marginable value of your (all cash) portfolio.
i.e., you couldn’t use the marginable value of that cash for something else while you have those short puts open.
But you’re still not borrowing any money, and not incurring a margin loan, so there is no cost for it.

If you write a naked put WITHOUT a cash pile to back it up, well, that’s a horse of different colour.
Since nobody sane does that, it’s not worth calculating the margin requirements of it.

Jim

2 Likes