A Cautionary Tale : To Trade or not to Trade
... there are more things in heaven and earth (or the markets), horatio than are dreamt of in your philosophy (or even your algorithmic trading toolkit software)
The grand test of the StockWave "meta-algorithm" cannot be done - the parameter space is too large, too high dimensional, and with too many stocks, data selections and choices to be made - the resources and the time it would all need is simply unavailable - and so no statistically significant claims can be made, and yet, there still needs to be some testing done, a base validation or to put it more crassly, to "eat one's own dogfood ..."
I decide to trade ten stocks; I will use basic models and look for trade combinations with good payoffs using StockWave and will then use the TradeMonster papertrading account with CBOE to place the trades and see how it goes, if for no other reason than to make sure that everything works and it all looks reasonable.
To start with I make some basic mistakes with the paper CBOE trading platform - I enter in some dumb trades that can't possibly make any money - "reverse arbitrage" (- the prices used by StockWave came from Yahoo and did not agree with the ones on TradeMonster) - I also enter in the wrong parameters, or enter custom spreads i.e. my combination trades, as individuals, costing me more in brokerage fees ... but after a while I am getting the hang of it, and my performance screen is mostly in the green, but there are two nasty red lines for JPM and AAPL. Seeing red irks me - for although it doesn't mean much from the view of statistical validation, I want to see all my trades in the green, even if it's not by much.
The JPM trade was really annoying - but illustrates an important point - it was as I wanted to trade it, but very quickly after about a week was heavily in profit - the JPM underlying was volatile and it moved the right way; but I wasn't monitoring the situation properly - I had just entered it to see what would happen, but I hadn't made an exit plan ... I failed to realise that I had about 85% of my profit target in my hands but failed to close it out, then one day later the price shifted the other way and I was looking at a hefty loss. I realise I am talking like an angler who has lost a fish off his hook ... and indeed there are a lot of similarities.
The AAPL trade was the first trade I had tried to make, and it was just "wrong" - a dumb trade that couldn't make any money even theoretically - I hadn't used the TradeMonster tools to check with the StockWave Trade Creator output - these should be in rough, but not exact, agreement.
Having learned my lesson I decide to remedy the situation - JPM first - I find myself a 3-leg combination, take a nice profit and move my overall figure for trading in JPM to be small but positive, in the green. I take a small satisfaction from the fact that my trade made (paper) money by effectively shorting JPM, who it must be said are one of the world's nastiest, most crooked and predatory financial organisations - their top boy, Jaime Dimon, a kind of "King Weasel" - if there is a scam or a manipulation on-the-go, they will be into it, so to take a few bucks out of them feels morally justified. Anyway, job done, next up AAPL - and will it be "jobs" done?!
AAPL should be a great stock to trade - a deep, liquid market, both in the stock and in the options; the options should be especially inviting as there is a vast range of strikes available and high open interest on almost all of them ... profits ahoy m'lads!
I hit a wee problem immediately - the Trade Searcher is using an exhaustive enumeration which works only up to 2 legs and crashes once one tries to do 3 - this is because the AAPL options chain has too many strikes; some hand editing is used to thin out the results, but I don't go as deep as I would like to - in the end I just look at the 2 leg combinations. This sort of situation really needs one to shift to using the optimising search algorithm, but it isn't ready yet and has not been tested (- it's quite a cute little thing, the first time I've ever used a genetic algorithm on a real problem - "geek" kudos and all that.)
I am just looking for a quick trade to bang-in and get myself into profit on AAPL - up comes a real top pick, a very high probability trade with a nice profit level ... the trouble is, it's a strangle and I remember my own written advice about trading combinations and about "red triangles" - trading a strangle is a bit like being in a room with two rapists - you risk getting rammed from both ends! Still, I can't see anything else to trade and it has 90% chance of profit, with 87% chance of maximum profit, the underlying is at that time 585 and the payoff range is just massive - from 522 to 662 at expiry I will be in profit, mostly maximum profit. This is a 10% price range either way, surely one can trade a strangle when the central "safe" zone is so utterly wide? There is 3 weeks or so to expiry, so I have a long time to wait, but in the end I just make the trade.
The thing about such a trade is that to start with it is heavily negative, any profit will come late in the day, or if you like the greeks, the theta is very high - you just need to sit there, letting the negative fade away until very close to expiry when you should be in profit.
For a little while nothing much happens, then the shit hits the fan - AAPL, the biggest, best, most loved and feared tech company in the US suffers its biggest one day fall in years - the markets are in uproar, the story is rippling across the internet, but without any convincing explanation - typically one might expect a "trendbreaker" news event to have caused this but there is nothing ... previous news is the iphone 5 is just out, the app store has opened in china, nothing much, christmas is coming and they should be selling tons of their gear ... what the hell is going on? Some weak explanatory offerings appear - some analyst wrote a research note that "android tablets are good too, you know" (- I know, I've got two ... is this news to anyone?) and something technical about clearing arrangements. I just don't buy it - AAPL is the biggest company in the US, and this is the best market news journalism has to offer on it? You would be as well-off reading the daily horoscopes.
One thing I have not been doing properly in this informal validation phase is using all features of my software - I am being a little sloppy and not following my own rules; I have not done any news analysis or even checked the news files for AAPL, but there is just nothing there anyway ...
Still, I am not that worried about this - the underlying is still in the safe zone, there is a lot of bouncing, but even with the downward trend - should still be OK. I feel this actually validates my approach - I have crafted a trade with a payoff zone so large that even the most extreme, unexplainable moves, cannot affect me. Clever boy!
Looking closely at the share price I notice something, a phenomenon I had not seen before, and one which continues and indeed gets worse all the way through the trade - the AAPL share price "jumps" or "gaps" at the opening, almost every single day - sometimes up, but mostly down, and it's always at the open, and there are no jumps or "trendbreakers" through the day. Most odd - for if you stitch the price in your head, eliminating the opening jumps the price dynamics look consistent, normal ... but these gaps look crazy, and it's always at the opening. My 522 lower level is crashed through, and the trade is looking in peril - the position is highly negative, but I let it ride, just hoping to let it recover. Note that this was an uncomfortable state when paper trading, I would not like to have real cash riding on it. Some internet loudmouths are calling 470 as the new price target ... thanks for that - not helping!
The trade recovers a little, hovering about 522 with a day to go, then on the last day crashes down, again, ripping away my profits - like someone has deliberately "taken me out" (- but that would be "paranoid", wouldn't it?) - the price recovers a little to hit 520 at the close on expiration day, a lousy 2 bucks away from my profit. Shit! Merde ! Schiesse ! I must admit that was an uncomfortable roller-coaster ride, and at the end I feel a little sick, and even ... "cheated", you see ...
AAPL is not some "dodgy" mining company, operating in a remote part of the world, with dodgy people and dodgy regimes, making dodgy claims of dodgy discoveries, listed on some "wild west" barely regulated exchange for small companies - it should be the most "known" and scrutinised company on earth, and yet there is something "funny" about its share price dynamics. Big moves, always at the open, never anywhere else, and no real explanations as to why - the news reports are not explanatory or believably causal, they are simply tautological - "big falls in AAPL" - reflective of the share price; there is something here that is unmeasured and therefore unknowable and cannot be factored into even our probabilistic notions - one must always be aware of this possibility, that there is something that we have not covered, it is like the dark matter of the financial universe, we know its there, but we cannot detect it directly by any conventional means.
What do I think myself? I think a lot of big players hold AAPL and hold related positions in various derivative contracts, I think when trading stops on the exchanges, a lot of guys, hedge fund managers, traders, bank principals, start phoning their counterparts, or even meeting in secret ... doing various off the record deals and horse trading - to protect their own arses, to make more money, coordinating themselves to stitch up the dumb money; of course this is probably very illegal, which is to say ... par for the course ... I sold a whole bunch of contracts at 525 I need taken out ... we need it to touch 515 ... I need the AAPL-GOOG pair to widen 10 points ... etc
Still, I won't do the "weeping virgin" routine, dabbing at a single tear on my cheek with a lace handkerchief, an innocent ravished by the unscrupulous and the predatory - I knew what I was doing and I admit I made some mistakes -
- I didn't follow my own advice about strangles; like some southern preacher praying to Jeeezus - praise the lord - talkin' 'bout SINNNAH and CORRRUUUUPTIONNAAAHHH ... but getting caught out in a motel with a hooker and an ashtray full of coke ... "saying it" isn't good enough, you have to believe it, you need to do it, you need to obey it.
- I was too eager to trade and also, greedy; there's always another day, another trade - if you are not happy, don't trade. I traded before I checked out the deeper combinations - there was probably something good and a lot safer if I was just more patient.
- I should improve the trade searcher algorithm, in particular, allow for optimising trades and also to restrict the results by trade type, to be able to say - "optimise win probability up to 4 legs deep, but don't give me any strangles"
- I didn't use all the features of my own software. Duh!
- In particular, I didn't have the alarm system setup. Double Duh!
- Always be aware of new phenomena, things you never suspected; this is one place where human intelligence trumps AI every time, to step back and say "does that make sense" or "something fishy here"!
- Don't make assumptions, in particular just because a stock is very big, the market deep and liquid, does not mean that someone, somewhere cannot cause unusual movements in it; when there's lots of money at stake, anything can happen, especially on a triple witching day.
- I had got a little complacent; I had been bashing out trades with small or decent profits without too much trouble - not concentrating like I should have - the unwatched cur will bite you in the arse.
- Trades where you are committed from the outset to hold it open until very near the end, are not good for the nerves.
- When it was getting close to, as Alex Ferguson would say, "squeaky bum time", late in the trade, hovering on the boundary of profit / loss I should have run a few advanced analyzers to see if I could get some directional or turning point information ... this may have allowed me a hedge, even very late in the day.
- Whatever happens, I won't trade AAPL again - ever - unless the NSA can provide me with the cellphone intercepts for wall street ... like that's going to happen!
- Watching the AAPL price moving down, it seemed to have an inevitability about it, like a collective will had decided to move it down there - just so - had AAPL announced a combined time machine and orbiting death ray with augmented reality goggles and a flying car all for a combined price of $99 ... or all of the previous plus warp drive for $129 ... the share price would still have gone down.
I tell you all my mistakes so you don't make them! Even having written StockWave, when it comes to actually using it I am really still a "noob" ... (- compare this to say, Karsten Solheim, the guy who makes the Ping golf clubs ... he was not a champion golfer himself, or even I think a low handicapper.)
A curious postscript - the trade was closed out on expiry automatically by TradeMonster - but something was wrong - it has credited me with having made an absolutely huge profit on this lousy trade - I check the contract ids and trade actions and they seem to be right, but the closing prices are very wrong ... which just goes to show how tricky all this business can be, if the brokers themselves get the details wrong! Of course, this is one mistake I might forget to tell them about ...!
