Shrinkage & LEAPS

Shrinkage is one of the better SIPro screens. It looks for companies that have been buying back a lot of their stock and then sorts by high ROE. Jamie’s backtester shows that, since discovery in 2003, for an annual hold the CAGR is around 13%, couple of points better than the S&P.

It seemed to be worthwhile to see how the screen has done when used with 12-month calls (LEAPS). The following results are based buying at-the-money calls after the third Friday in January. The three positions out of the screen’s 10 picks that had the lowest implied volatility (cheapest) were chosen. Of course, not all of the stocks had LEAPS or even options particularly in the early years.

I looked at two approaches:

  1. buying calls with 24 months until expiration and holding 12 months
  2. buying calls with 12 months to go and holding until expiration

For comparison I’ve included results for LEAPS using the Plow26wk screen.


Start               Average return
Date     Plow26    Shrinkage 24/12   Shrinkage 12/0
1/21      100%          63%               71%
1/20      245          160               262
1/19      148          268               425
1/18     - 59         - 65              -100
1/17      517          576               902
1/16       61           83               194
1/15     - 34         - 95              -100
.
Average   140%         141%              236%

Here are the rest of the Shrinkage post-discovery numbers using the 12/0 approach:


1/14    - 17%
1/13     598
1/12    - 22
1/11      42
1/10    - 59
1/09      38
1/08    - 69
1/07    -100
1/06    -  6
1/05      30
1/04      10
.
18-yr average:  115%

DB2

11 Likes

At the money LEAPS might be the most expensive way to do it…the highest time value, so likely the highest drag from decaying time value.
I would speculate that slightly in the money would do better, maybe more than slightly.
The losing positions will lose more, but all the positions will start off at a better breakeven.

12 month at-the-money LEAPS are pretty expensive.
For your 12/0 strategy:
I’m not sure what companies the screen picks, but some likely biggies might be Apple or Oracle.
High ROE, lots of buybacks, anyway.
An 11 month first-out-of-the-money call on either one costs about 12% of the stock price right now.
That’s quite a drag to overcome.

Jim

7 Likes

At the money LEAPS might be the most expensive way to do it…the highest time value, so likely the highest drag from decaying time value. I would speculate that slightly in the money would do better, maybe more than slightly.

I’ll check it out.

DB2

4 Likes

At the money LEAPS might be the most expensive way to do it…the highest time value, so likely the highest drag from decaying time value. I would speculate that slightly in the money would do better, maybe more than slightly.

There is some improvement in pricing, but the leverage effect dominates. In down years you do better (less poorly) with in-the-money calls.

Here are the results for the last 14 January cycles:


Start     1st    1st    2nd
Date      OTM    ITM    ITM
1/21       71%    83%    88%
1/20      262    217    193
1/19      425    342    300
1/18     -100   - 86   - 76
1/17      902    730    598
1/16      194    167    155
1/15     -100   -100   -100
1/14     - 17     21     35
1/13      598    432    363
1/12     - 22   -  1     16
1/11       42     55     61
1/10     - 59   -  7     16
1/09       38     70     65
1/08     - 69   - 58   - 36
.
Average   154%   133%   120%

DB2

7 Likes

There is some improvement in pricing, but the leverage effect dominates. In down years you do better (less poorly) with in-the-money calls.

Here are the results for the last 14 January cycles:

Start 1st 1st 2nd
Date OTM ITM ITM
1/21 71% 83% 88%
1/20 262 217 193
1/19 425 342 300
1/18 -100 - 86 - 76
1/17 902 730 598
1/16 194 167 155
1/15 -100 -100 -100
1/14 - 17 21 35
1/13 598 432 363
1/12 - 22 - 1 16
1/11 42 55 61
1/10 - 59 - 7 16
1/09 38 70 65
1/08 - 69 - 58 - 36
.
Average 154% 133% 120%

DB2

Thanks so much for sharing these results!

The averages you show seem to indicate that the 1st OTM Strike Price is the clear winner but this deserves some more thought.

As has been discussed here before, nobody invests their entire principal in such a strategy, but the actual allocation does make a difference.

As an example, let’s assume we begin 2008 with a principal of $100 and allocate 30% ($30) of it to the 1st_OTM strategy, leaving 70% ($70) as cash (earning 0%). In the first year, we lose 69% of the $30 and none of the cash, leaving us with $79.30 to start 2009. The 30% allocation resulted in an effective total return of -21.7%.

The next year, we put 30% of this $79.30 into the strategy ($23.79) and keep the rest in cash ($55.51). By January of 2010, we have a total of $88.34. Here’s a general description of the Total at the start of each year (t).


     Total[t] <- Total[t - 1]*(alloc*(strategy_return[t - 1] + 1) + (1 - alloc))

Rebalancing each year to 30% in the strategy and 70% in cash, the effective returns for the table you provided would be:


  Start OTM1st      Total
2008-01    -69% $100.00000
2009-01     38    79.30000
2010-01    -59    88.34020
2011-01     42    72.70398
2012-01    -22    81.86469
2013-01    598    76.46162
2014-01    -17   213.63376
2015-01   -100   202.73844
2016-01    194   141.91691
2017-01    902   224.51255
2018-01   -100   832.04349
2019-01    425   582.43045
2020-01    262  1325.02926
2021-01     71  2366.50226
2022-01     NA  2870.56724

The Total CAGR of this strategy (OTM_1st with 30% allocation) is 27.1% ( i.e. (2870.57/100)^(1/14) - 1 ). One of the reasons many people think of things in terms of log-returns (the log-return is: ln(P1/P0) and the arithmetic return is: (P1 - P0)/P0 ) is that we can also arrive at this CAGR by taking the average of the log-returns and then converting it to arithmetic return using: arithmetic return = exp(log-return) - 1 ).

Rebalancing to different allocation fractions will result in different CAGRs. Here’s a table with various allocation fractions (0 to 100%) simulated on the 14 returns given for each of the three strategies tested and posted by DrBob2:


alloc   OTM1st     ITM1st     ITM2nd
  0%   0.00000%   0.00000%   0.00000%
 10   12.45532   11.37782   10.62093
 20   20.99829   20.18077   19.29307
 30   27.11807   27.10045   26.47019
 40   31.28839   32.71759   32.47637
 50   33.54719   36.57261   37.29168
 60   34.04351   39.33219   40.96422
 70   32.16856   40.42181   43.13471
 80   27.67465   38.79042   43.10850
 90   17.38093   33.38359   39.74313
100 -100.00000 -100.00000 -100.00000

Notice that for an allocation of 0%, the return will be 0% (i.e. same as Cash) and for an allocation of 100%, the return will be -100% (i.e. at some point we would have lost everything).

These data simulate what would have happened if we had invested and rebalanced as described and the returns are as reported. If we can assume that the yearly returns will be distributed in the same way for any year in the future, then we can get a better idea of the distribution of possible CAGRs by using taking repeated samples (with replacement) for each case from these 14 years of data (this is bootstraping).

Using this technique (w/ 2,000,000 re-samples for each strategy), the optimal allocation for each strategy is:

OTM1st optimal alloc is: 57.31% with a CAGR of 33.98%
ITM1st optimal alloc is: 69.65% with a CAGR of 40.25%
ITM2nd optimal alloc is: 75.74% with a CAGR of 43.53%

To summarize, it’s not clear that the average of the arithmetic returns gives enough information about which strategy to use. The allocation/rebalancing strategy also needs to be considered.

heink

15 Likes

To summarize, it’s not clear that the average of the arithmetic returns gives enough information about which strategy to use. The allocation/rebalancing strategy also needs to be considered.

Very true. Allocation, rebalancing and portfolio management were left as an exercise for the reader.

My personal preference with option strategies is a fixed allocation each year with no compounding with lots of backup for the losing years. Then just let the strategy throw off cash over time (presumably) which can be used to enjoy life or to try other strategies.

Thank you, heink.

DB2

3 Likes

As has been discussed here before, nobody invests their entire principal in such a strategy, but the actual allocation does make a difference.

To summarize, it’s not clear that the average of the arithmetic returns gives enough information about which strategy to use. The allocation/rebalancing strategy also needs to be considered.

Excellent point. And great work.

Though there are allocation strategies other than the ones you tested, of course.

If it’s a “corner” of a larger portfolio, one might simply have it a constant size every year: rebalance the strategies.
i.e., after a loss, top up the options segment from other sources, rather than having a smaller bet after a bad year for this strategy.
And, presumably take profits out of the strategy after particularly good years.
In this situation, it doesn’t need a companion cash counterweight, assuming it’s a small enough
portion of the overall portfolio that there won’t be a problem topping up the cash from savings or elsewhere.

In that case, the simple average of annual returns is probably a better guide.
And at least easier to calculate : )

Jim

5 Likes

In that case, the simple average of annual returns is probably a better guide.
And at least easier to calculate : )

Amen, brother.

DB2

I appreciate your analysis on this. I rarely post here, but use options heavily in my portfolio, usually as far out in time as I can get, using an LTBH point of view.

Another consideration is that with options you can roll down or down and out when the market drops significantly. On a drop like this I have added money to roll down further (and out), as I expect high quality stocks to recover within 2 years. This can result in significantly juiced returns.

And as noted by the heink, allocation is the part of using option leverage that is critical and must be managed carefully. Anyone without much option experience should start small.

Enjoy,
Brian

3 Likes

Another consideration is that with options you can roll down or down and out when the market drops significantly. On a drop like this I have added money to roll down further (and out), as I expect high quality stocks to recover within 2 years.

Do you have some general guidelines? Do you take any money off the table?

DB2

DB2,

You’ve discussed some of this (your allocation style) in previous posts. True, the averaging of conventional returns does get you the “right” answer for small amounts invested (relative to your total). I just didn’t really trust that answer until I actually went through the simulation exercise.

I very much appreciate the options backtesting strategies and data you’ve shared and at some point might want to join you.

Options strategies involve asymmetric returns (e.g., a high probability of a small loss balancing a small probability of a huge gain) and it’s not easy to understand what the actual risks and returns are. Because of this, such strategies might stand a better chance of providing a little pocket of inefficiency that can be exploited by a properly back-tested mechanical approach.

heink

In that case, the simple average of annual returns is probably a better guide.
And at least easier to calculate : )

Jim

Since early in my MI days, I’ve done everything using log-returns (I did a lot of RRS). I wanted to better understand DrBob’s backtests and this exercise was as much a way for me to to handle the presence of -100% returns (which is a log-ret of -Inf) as anything else.

heink

Do you have some general guidelines? Do you take any money off the table?

Rolling down is an art. It’s best with stocks you know really well and have a feel for when it’s not going down much more. If it drops lower and you roll down again, you’ve mostly wasted the premium paid for the first roll. To get a feel for it, use small positions and take your time. It’s good to keep a journal of the reasons for every purchase, roll, or sell so that you can track how well those reasons work in the long term.

Since options expire, you either take money off the table by selling or rolling and pay any taxes, or you exercise to stock. The advantage of the latter is that you don’t take taxable profit, but you must inject cash to cover the shares bought at the strike price.

Enjoy,
Brian