Crash and burn

Screening for stocks that will crash and burn is not an easy task. Our short screens pick those that have already crashed. We are in on the burn phase and that can go either way. Witness the success of the NegFCF screen for buying calls.

How does one identify stocks that are going to crash in the next month or two? The Altman_Z approach is more oriented toward an annual timescale. If someone has some ideas, perhaps from technical analysis or volatility studies, please speak up.

We do have an MI that looks for slowing momentum, FOG_MI. Flare Out Growth was introduced by Jon Markman and is explained in this post from 1998:
https://discussion.fool.com/flare-out-growth-10452385.aspx
It looks for high momentum stocks that have paused. But what if a significant number of them don’t resume their upward trajectory?

Here we buy at-the-money puts on the first three picks of the FOG_MI screen. They have two months until expiration and are held for the two months. Here are the results – no compounding – for the last four years. (Note: since 1/22/18 SPY is up some 67%.)


2021    132%
2020      4
2019  -  24
2018   1092

Even the bad years are not that bad, and in the one year that the S&P was down (2018, -4%) the puts hit it out of the ball park.

I can post the returns for each cycle if there is interest.

DB2

15 Likes

Each month one could buy puts on three FOG stocks and calls on three NegFCF picks. Here is how that would have worked in 2018:


Start
Month    Puts   Calls
18-01      6%   - 55%
18-02     14    -100 
18-03     16      94
18-04   - 26     451
18-05      2    - 33
18-06     73     269
18-07   - 29      57
18-08    264    -100
18-09    434    - 45
18-10    481    -100
18-11   - 81    - 23
18-12   - 55     363

DB2

8 Likes

An idea:

If you wanted to cull the picks a bit to try to find the most likely failures, might I suggest as a final sort: lowest ratio cash/market cap.
Cash rich firms tend to be more resilient in the fact of adversity.
And so, presumably, the reverse.

The screen definition also requires Timeliness=1, if I recall.
If you skip that, long run returns drop by 3%/year and average loss widens.

So a minimal version of the screen could be:
tr1y - r13 - 3 * tr4w top 10
csh / mcp bottom 5

Average loss per pick among the losers since 2000 in a two month hold -18%, and more losers than winners.
Two of the top 5 picks from Jan 3 have dropped nicely year to date, THRY and CROX.

Not such a terrible short screen, in general.
5 picks 2 months no friction underperformed SPY by about 13%/year 2000-2021 inclusive.
Unlike so many short screens, it didn’t return a gazillion percent (killing the shorts) on rebounds from plunges in 2009 and 2021.

Jim

10 Likes

Unlike so many short screens, it didn’t return a gazillion percent (killing the shorts) on rebounds from plunges in 2009 and 2021.

I like well behaved short screens!

Thanks for the idea.

DB2

2 Likes

From the charts - - Maybe a moving average for an individual stock whose slope has become a certain value less than its sector and its rate of deviation away from such sector trend is increasing. Look for option activity favoring a demise and then some fundamental trends for icing on the cake?

d(crashing)/dT

1 Like

Maybe a moving average for an individual stock whose slope has become a certain value less than its sector and its rate of deviation away from such sector trend is increasing.

I was also wondering if there might be a change in volatility or the standard deviation of the trend.

DB2

1 Like