Hedged options selling systems sys14 and sys15 by botpro

botpro

Guest
Messages
21
Likes
0
Hello, I'm new here.
I have recently developed the following options selling system.
It has not been tested in the real markets yet.
Any comments, feedback and ideas to further improve it is kindly welcome. Thx.
Code:
Hedged options selling system sys14 (and the combo-system sys15)
AN OPTIONS SELLING SYSTEM THAT THEORETICALLY ALWAYS WINS THE FULL CREDIT

Author : botpro at [url]www.trade2win.com[/url] (previously botpro at [url]www.elitetrader.com[/url])
History:
  2016-04-19-Tu v1.4a Added legal stuff
  2016-04-18-Mo v1.4  It works also with American style options
  2016-04-17-Su v1.3  Added optional Extension #1 (s. end of the text)
  2016-04-17-Su v1.2  Fixed PnL to 201% p.a.
  2016-04-17-Su v1.1  Fixed PnL to 218% p.a.
  2016-04-17-Su v1.0  Initial version: PnL=187% p.a.


First things first: the legal stuff:
  To be on the safe side, I as the author of these free systems make these statements:
  1) The systems are the result of academic research.
  2) The systems have not been tested yet in the real market.
  3) Use the systems at your own risk. Test it first throughly in simulations and/or backtestings/forwardtestings.


Intro:
  - This is an options selling system with a special hedging mechanism (called "entry-level hedging").
  - A classic static or dynamic zero-delta-hedging is not used with these systems.
  - One needs a near-24/7 market with ideally no overnight gaps in the underlying stock price.
    Such a market can be emulated by using multiple markets around the globe where the stock trades.
  - The hedging mechanism requires that the stock position gets monitored continuously.
  - If applied correctly in the market, then it is a zero-risk system.
  - It can be used for all options regardless of the option style (American or European).
  - PnL is dependent on the ATM premium: the higher the ATM premium the higher the PnL.
  - The basic system uses a short Call (can also be a short Put, but then hedging rules have to be reversed).
  - The combo-system (sys15) uses a short Call and a short Put together (but uses just one stock position for both!).
  - The example uses a 40% volatile stock; then the PnL is about 201% p.a. for sys14 and 658% p.a. for sys15.
  - It is the special hedging mechanism that makes it all possible.


Beware:
  - The example below uses for simplicity just the minimum 1 contract,
    but in practice one would need a PDT-account with at least $25k.
  - When testing this system use a cheap stock commission rate like that of the broker IB ($0.005 per share, minimum $1).


Rules of the options selling system sys14 with "entry-level hedging":
  1) The system consist of 2 parts:
       1.1) sell a call option, and
       1.2) go long the stock
  2) The option position is normally kept till expiration (but a possible early assignment is OK with this system)
  3) Entry-level hedging for sys14:
       3.1) if stockprice drops below the initial entry price then close the stock position immediately
       3.2) if stockprice crosses over from below the initial entry price then re-open the stock position immediately
       In practice one would use two bands (for example +10 cents and -10 cents) around the entry price to avoid too many hedging trades.
       (FYI: the hedging mechanism for sys15 is a little bit complicated and tricky; in the next version it will be documented)


Example: assuming InitialStockPrice=100, HistoricalStockVolatility=40%, using monthly options (ie. Expiration=21 business days)
  Selling 1 Call:
    Spot=100.00 Strike=100.00 ExpDays=21 HoldDays=21 EarningsYield=0.00% DividendYield=0.00% StockVolatility=40.00% --> Call=4.60 Put=4.60

  Going long 100 shares of the stock:
    Spot=$100.00 --> CapitalNeeded=$10,000

  At expiration date:
    Regardless what the CurrentStockPrice or the CurrentVolatility is: we will keep the credit fully.
    The stock position serves us only to hedge our option position; it is not intended for making additional profits.

  Profit:
    We use the usual 2:1 overnight margin of the broker. So we have a leverage factor of 2.
    Meaning: of the $10,000 only half of it minus credit / 2 is our own investment (this is the basis for the PnL calcs).
    CreditReceived = $4.60   * 100                 = $460
    StockPos       = $100.00 * 100                 = $10,000
    UserPart       = $10,000 / 2 - $460 / 2        = $4770  (this is the basis for the PnL calcs)
    MonthlyPnL     = $460 / $4770 * 100            = 9.64%
    AnnualPnL      = ((1 + 9.64/100)^12 - 1) * 100 = 201.73%
    From these numbers the commissions paid and the interest for the margin has to be subtracted.


Remarks:
  - It's up to you to apply this system in the market(s).
  - To get the hedging correct, one would need to monitor and if necessary trade the stock nearly 24/7 by
    using multiple exchanges around the globe. Ie. by this, one has to eliminate any overnight gaps in the stock price.
  - If applied correctly then the given guarantee holds, ie. in this example 201% profit per year.
  - Using more volatile stocks and/or a shorter timeframe than the above used 1 month will give even more profit.
  - Regarding margin: you have to do the math to get the 2:1 margin. Ie. in this case you would initially
    keep only $5000-$460/2=$4770 plus commission in your account. The broker grants you 2:1 overnight margin (or 4:1 intraday margin).
  - The system is freely scalable to any investment amount. But the stock position should not be too big
    because it must be easily closable and re-openable; I would say the stock position should not be more than $50k.
  - For big money one would use multiples of such constructs, but then one should of course use different stocks.




The following extension(s) are optional:
--------------------------------------

Extension #1: Profit booster: How to make more than 8000% per year with this system:
  The smaller your own part of the invested total money is, the more the profit% will make up:
  Example:
    - let's say your own money is $1000
    - borrow the rest ($3770) cheaply from friends or take a 1 month loan
    - let's assume you have to pay 1% interest for the 1 month loan, ie. $37.70
    Now, your basis for the PnL calcs is your own $1000 plus the interest for the loan = $1037.70:
      MonthlyPnL = 460 / 1037.70 * 100              = 44.3288%
      AnnualPnL  = ((1 + 44.3288/100)^12 - 1) * 100 = 8070.26%
    Since compounding gets used, each month the above absolute numbers would of course change,
    but the relations (ie. the percentages) would stay the same.


--- end of text ---
 
Last edited:
Hmm...

You're not really hedging, since you cover 100% while the ATM call has a delta of 50.
Try 50% hedge, so a move in the stock is actually offset by the option.

Second, your'e also playing a volatility trade. Vol up, you lose...

Third, in a volatile market you'll be wiped out. You underestimate the number of trades in the stock you need to do and the constant losses you will generate with that. Especially when you cut at 10cts up and down (that's 0.1%).

The active delta hedging your'e trying to do, should be linked at the actual delta your total position has. The higher the Implied Vol should mean you can let the stock move a bit further away.

Anyway, good luck with 24/7 markets... there's no way you can get that sorted. Even if trading the underlying stock on an overseas market in the after hours, you will eventually have to trade out.
 
Hmm...

You're not really hedging, since you cover 100% while the ATM call has a delta of 50.
Try 50% hedge, so a move in the stock is actually offset by the option.
Thank you for your analysis.
As clearly stated in the text, this system does not use delta-hedging; instead it uses a new kind of hedging called "entry-level hedging".
So, the delta plays no role here at all.

Second, your'e also playing a volatility trade. Vol up, you lose...
Hmm. how come? The hedging is done with the underlying stock only. The option position is constant. Then vola changes should not play a role with this system, IMO.

Third, in a volatile market you'll be wiped out. You underestimate the number of trades in the stock you need to do and the constant losses you will generate with that. Especially when you cut at 10cts up and down (that's 0.1%).
It is only at the entry price level. If the current price is away then it doesnt't have any impact on the system.
I think the too-many-crossings-problem is a solvable problem.

The active delta hedging your'e trying to do, should be linked at the actual delta your total position has. The higher the Implied Vol should mean you can let the stock move a bit further away.
As said, delta-hedging plays no role in this system. It uses a different, a new hedging mechanism. I think it is even better than delta-hedging.
The strategy is similar to a covered call, but is not the same.

Anyway, good luck with 24/7 markets... there's no way you can get that sorted. Even if trading the underlying stock on an overseas market in the after hours, you will eventually have to trade out.
Yes, that is the weak point. But OTOH that problem of overnight gaps is given for all hedging methods; it's a common problem of all hedging methods.
 
Last edited:
Thank you for your analysis.

You're welcome.

I guess it could work when you're in time to 'hedge' and sell your stock when the price drops, probably with a stop?
In the end, this will be a delta trade. If you're in time to sell your stock, you'll be delta short through short call. If the stock moves up, you will be net delta long.
So you're aiming to get the premium...

Why would you start by also buying the stock? You might as well start by selling the call naked, and do your hedge when the stock moves. The difference will be you start delta short and pay less fees because you're not buying the stock yet.

By Vol I mean implied volatility of the options. Which is basically controlled by the market makers. If they decide the this should be higher, the option value will be higher, no matter where underlying is.

But, I see this as a delta trade with short gamma, which will cost you dearly if you're too late when markets move rapidly.
 
Jump-diffusion for GBM

Today I did test the system using GBM simulations with normal and leptokurtic distributions (the latter has slightly "fatter tails").
The result looks IMO good: on average about 30 hedging trades per month (of course the less this number the better).

But I have to add jump-diffusion into GBM to make it more realistic by also including gaps into the model, ie. simulating a noncontinuous process.
Currently working on this problem.

Would be nice if someone could point me to source code (preferably C/C++) for jump-diffusion, to be added into GBM,
or a GBM that already includes jump-diffusion.
I think the basic jump-diffusion model of Merton (1976) should be sufficient for my needs.

Thx
 
Last edited:
Just an idea, what if you do a trade to end delta neutral at the end of trading, to minimize risk at a gap opening... That will mainly be around option strike price...
Reopen your strategy at next opening...
 
Just an idea, what if you do a trade to end delta neutral at the end of trading, to minimize risk at a gap opening... That will mainly be around option strike price...
Reopen your strategy at next opening...

An interessting idea, indeed, thx; it surely deserves testing in some simulations and/or with real data.
 
Update:
I tried the system now also with a q&d (ie. non-scientific) jump-diffusion emulation (10 jumps per day, normally distributed, taken simply pctChg = random value from normal-dist (ie. about -4..+4) :): it gives on average 31 trades per month.
This is IMO still a good result.

Update-2:
Even if I multiply the above pctChg by 2.0, the result is still about the same. That's IMO very good!

...because statistically half of the total number of gaps occuring are an advantage for the open position... ;-)
ie. on average it squares (meaning it equalizes)...

A true jump-diffusion implementation is on the way...

.
 
Last edited:
Basically your system looks like a copy of this....

http://capturingtheta.com/

Someone else has thought of this already.

I have tested this to death and IT DOES NOT WORK. No matter how small or large you make the gap from when you close a hedge to when you put it back on again - it bleeds money from your account and you end up losing over time. Some months you win big especially if the market charges in the direction of the hedge - but a lot of the time the market just meanders about a single point - and that's when you lose lots of money closing a hedge and then putting it back on again.

My advice is find something else.
 
Code:
Jump/Gap Stats: nRuns=100  (each run is 21 days, ie. a month):
  Totals: cJumps=21000.0 :
     -10.0%     0.0
      -9.0%     0.0
      -8.0%     5.0
      -7.0%    25.0
      -6.0%   107.0
      -5.0%   333.0
      -4.0%   900.0
      -3.0%  1876.0
      -2.0%  3195.0
      -1.0%  4087.0
       0.0%  4001.0
       1.0%  3093.0
       2.0%  1974.0
       3.0%   932.0
       4.0%   335.0
       5.0%   108.0
       6.0%    23.0
       7.0%     6.0
       8.0%     0.0
       9.0%     0.0
      10.0%     0.0
  Averaged: cJumps=210.0 :
     -10.0%     0.0
      -9.0%     0.0
      -8.0%     0.1
      -7.0%     0.2
      -6.0%     1.1
      -5.0%     3.3
      -4.0%     9.0
      -3.0%    18.8
      -2.0%    31.9
      -1.0%    40.9
       0.0%    40.0
       1.0%    30.9
       2.0%    19.7
       3.0%     9.3
       4.0%     3.4
       5.0%     1.1
       6.0%     0.2
       7.0%     0.1
       8.0%     0.0
       9.0%     0.0
      10.0%     0.0

TotAvg: cHedgingTradesPerMonth=28.1


.


.
 
Last edited:
Hmm...

You're not really hedging, since you cover 100% while the ATM call has a delta of 50.
Try 50% hedge, so a move in the stock is actually offset by the option.

Second, your'e also playing a volatility trade. Vol up, you lose...

Third, in a volatile market you'll be wiped out. You underestimate the number of trades in the stock you need to do and the constant losses you will generate with that. Especially when you cut at 10cts up and down (that's 0.1%).

The active delta hedging your'e trying to do, should be linked at the actual delta your total position has. The higher the Implied Vol should mean you can let the stock move a bit further away.

Anyway, good luck with 24/7 markets... there's no way you can get that sorted. Even if trading the underlying stock on an overseas market in the after hours, you will eventually have to trade out.

.....what he said.;)
 
Basically your system looks like a copy of this....
http://capturingtheta.com/
Come on, my system is much different.
BTW: please be careful when you link such IMO scam-sites... It's an insult to the intellect what the guy writes there
(for example he writes "There is a 100% probability of an option losing 100% of its time value" :clap: )
ROTFL! It is not even wrong!... :LOL:

Someone else has thought of this already.

I have tested this to death and IT DOES NOT WORK. No matter how small or large you make the gap from when you close a hedge to when you put it back on again - it bleeds money from your account and you end up losing over time. Some months you win big especially if the market charges in the direction of the hedge - but a lot of the time the market just meanders about a single point - and that's when you lose lots of money closing a hedge and then putting it back on again.

My advice is find something else.

I did some improvements to the rules in the OP, and it works fine.
In a few hours I'll post the changes.
 
Last edited:
Yes.....you will find that it Works better statistically for what your hoping to achieve,
 
Yes.....you will find that it Works better statistically for what your hoping to achieve,
Do you know statistics?

Come on, sorry to say this but that is IMO BS. Because: the shorter the timeframe the better the annual PnL. So it doesn't make any sense to increase the period... Period! :cheesy:
 
Last edited:
Please don't post any off-topic crap here, only constructive and qualified contributions are welcome... Thx...
 
Good news: number of hedging trades per month reduced from about 30 on average to just 19 on average.

I think that's it.

Now I'm going to clean up the code and post the relevant part, ie. the new hedging logic of sys14.

<...this message has been updated several times...>

Update:
- delayed; requires some more time...
- it seems the original entry-level-hedging is still superior to other variants I tested.

.
 
Last edited:
Why do you limit your jumps to 7%? Are you planing to trade index only?
With single stocks you need at least a few jumps >15%...

I suggest you get some real-time data, ES-futures would be best. Trades almost around the clock with tight spread so quite reliable. Also easy to work out options values, either calculate your own on the basis of historical Implied Vols or using real option prices.

Markets are a lot more erratic then you think.
 
Top