iBankCoin
Joined Oct 26, 2011
153 Blog Posts

Correlation, Options And Kelly Criterion

Long series of posts coming up that will probably be quite full of words, but there is a lot that I really want to cover that I think will be a huge difference maker to many as a trader.

The typical market is made up of sectors that are pretty highly correlated with itself. Take for example, the S&P sector SPDRs, this is a correlation matrix.
correlationofmarket
You might notice that the average diversified portfolio is therefore around a .70 correlation over the last 6 month period tested.

The closer the number is to one, the more a strategy should change into

1)Picking individual stocks as opposed to broad diversification.

2)Concentrating a portfolio in fewer names across fewer sectors.

3)Having more cash on the sideline (to compensate for a strategy that is more levered to the market’s ups and downs as opposed to the individual stocks up and downs.)

4)Considering “alternative” investments

5)Considering a hedge and often also using less “leverage” overall.

Return is very directly associated with correlation. You will see this through discussion of what’s known as the “Kelly Criterion” which long ago I blogged a bit about. Basically a “kelly” is the optimal bet size for ONE bet with the rest in cash given a certain probability of winning and certain edge (upside vs downside).

For now we can just use past illustrations of long term growth based upon bet sizes as a percentage of the “kelly” to teach you about risk and reward.

Say a person who hates money or never learned statistics or common sense offered you a game. Tails you lose your entire bet, heads you win your bet back and 2 more bet sizes on top of it. 2:1 odds on a 50% chance of winning. He also has guaranteed that this “game” will be available indefinitely. But you can only play with the cash in your pocket and you can never go back and get more.This is basically the scenario considered with a standard kelly criterion calculation.

If you put all your eggs in one basket, you eventually go broke, even though your “expected value” of a given flip is very high. The reason is portfolio volatility. But if you risk too small you don’t get to maximize the gain either.

This problem is actually calculated and illustrated nicely. As is any sort of bet that offers an edge. The assumption is you have to bet a fixed percentage of your bankroll, rather than gambling, doubling down, etc.

The kelly criterion defines the “ideal” bet as the one that maximizes long term growth rate. There are so many reasons not to risk this much and to risk LESS than this amount.

I have since constructed my own modified Kelly Criterion calculator on a spreadsheet that can handle multiple inputs of various probabilities of various outcomes. In other words, it’s better than your average kelly criterion calculator. I may upload this spreadsheet at some point to share it with those interested so you can experiment with different data points and see how it effects your long term return.

I have went one step further and used my understanding of correlation to also allow inputs of MULTIPLE assets (assuming the same expectations of probabilities of various outcomes) with a given correlation rate. In other words, it allows for 2 or MORE bets into different assets simultaneously. After this application it actually becomes possible to get specific numbers for constructing a portfolio. If you want, we can get into the math sometime, but for now just a brief conceptual understanding.

At some point, I even want to take this farther for it to give you a probability of a given drawdown over a fixed amount of trades (such as 20% over 100 trades) using the strategy so it gives you a better snapshot as it relates to risk. I will have to run “permutations using macros” which basically means running all possible combinations that a series of 100 trades could play out, weighted by their probability of occurring. From there you can add up all events where the total bankroll as a result of those trades at given bet size adds up to a certain threshold lost. It will probably be a few months until I get around to getting the spreadsheet to that point since I have other things to tend to.

Ultimately regardless of how sound the kelly criterion calculator and strategy as it relates to risk management is, it means nothing if you don’t properly model your expectations… and of course have an edge to begin with. Simply using past results or anticipated results may over inflate the results and lead to overtrading based upon higher perceived results or overconfidence. So the most important thing is to understand that in an uncertain environment such as stock trading, less is more, and that has actually been proven mathematically.

Payout Heads=200%, tails=-100% probability of each event=50%. Full Kelly achieved at F%=25%

coinflip

You can see that you can actually maximize your bet size by betting the proper balance. The actual amount can be calculated in this scenario.  The full kelly or “ideal” bet to maximize long term growth rate per bet ends up being 25% of your bankroll. Maximizing long term growth is reckless in it’s own accord since that assumes you are around indefinitely to make up for wide volatility swings and maintain the composure to not capitulate and change strategies after. Anything beyond the kelly is insane because you don’t produce greater growth and incur much greater volatility and even risk of ruin if you go too far. If you bet a single cent beyond 2 times the “ideal” calculated amount you eventually reduce your bankroll to effectively nothing, it’s only a matter of TIME.

Additionally, you can calculate what your “wealth” growth rate is. Each bet would expect to have a long term growth rate of 6%, meaning your portfolio/bankroll GAINS 6% per bet on average. (Note:In the future, I have plans to use this to show you how to TARGET specific annualized returns based upon a given strategy.)

The spreadsheet itself will work with more than just individual “all or nothing” bets and can be structured for up to 16 outcomes (I could do more, but see no need). It also will work with multiple bets at a different correlation to tell you how much to risk per each individual bet.

In the next post I will go over how these concepts change as you use multiple bets correlated at less than 1, rather than just 1 all or nothing bet. In the future I will go into how you can use this understanding to optimize a portfolio for wealth accumulation, or target a return and minimize the risk while aiming for those expectations.

Comments »

The Moves

I sold my remaining SPY June 07 puts for a nice gain. I sold Yen calls for a nice gain. I reduced my LNKD puts for a nice gain. I sold my VXX calls for a very small gain. I still have my SOHU puts and TRV puts and ASTX puts… and my long term December SPY puts as cheap insurance (cheap on a cost per month basis, especially since I will roll them over or sell them well before the theta decay really hits.

For the time being I still own the dreadful metal called Gold but that probably won’t last. Copper is up yet FCX is kicking me today.

I scooped up some CAT weekly calls very close to the low and followed OA on One of his plays here… not saying which ones to protect the tremendous value his members get at trading addicts.
Update:I see he listed it in the comments, so I suppose it’s “safe” now to say it was AMZN.

Also took off my SODA calls today after following OA for a big gain.
I held too long on AFFY, but not selling at high prices isn’t a reason to sell at low ones. The structure is just fine it’s still making higher lows on a short term chart, and it’s still above it’s breakout price of $1.40. I am still long FRO of course, from $1.85

I am not just taking my shorts off to rebalance, I am shifting towards a bullish bias more aggressively, taking off some of my neutral plays (like FXY) and adding. However, that does not mean that I don’t have ways to make money if market goes down.

CASH is still a position of course, as always. I like the way commodities and oil are moving today. I like the way the market found a bottom off support, holding off the volume pocket.
I am still worried about volatility ahead and possibly some kind of dramatic crash, but that’s what my puts are for.

Comments »

Reality Check: Falling TLT (Rising Yields), Means What For Stock Prices?

I get a little tired of the myth that rising interest rates is bearish. Sure, periods of time it can be, but eventually the cash flooding out of treasuries goes somewhere. Also, rising interest rates mean there is a BID for CASH, so people are willing to start BORROWING. If people don’t want to borrow, interest rates don’t magically start going up. The interest rates go up on demand to bring cash up front and borrow.. Hence, this should be bullish if anything. Granted, the market behaves based off of ideas that were formed hundreds of years ago in economic textbooks. Sometimes falling bonds and rising yield has been bearish for stocks, but the long long term correlation and current trend seems to say rising yield is correlated with rising stock prices.

TLT going up means falling yield, rising bond prices. TLT going down means Rising yield falling bond prices.

TLT to SPY

correlation tracker

TLTnegativecorrelation

Sometimes you have to shut your economic textbooks, and look at reality.

Comments »

OK

So I continue to play both sides of the trade.  My FXY calls have been working nicely here, and I decided to add a real speculative June UUP $23 calls for a nickle today.  I wasn’t ready to sell my FXY but basically I am starting my rotation back out of the yen and into the dollar early. Remember, I entered the dollar and rolled some gains into the yen, so basically I am rotating it back except still holding the yen. The correlation between yen and dollar are not strong or inverted, unlike the dollar and euro which has a -.95 correlation.

As a result I can add and subtract at different times and both plays still provide negative correlation (but a very small one) to the market. This helps me keep my portfolio grounded when I get carried away buying a little too much.

I may exit my TRV put tomorrow if it doesn’t turn. I grabbed some SOHU puts today. So if I scratch out of TRV it doesn’t disrupt my exposure. If we trade higher, I close out of TRV and keep LNKD and SOHU on. If we close lower, then it is a bonus. My LNKD puts I am holding.

My GLD calls and FCX calls are still working well also. Had GLD not opened and closed above $135 today I would have scratched the trade. Lack of a follow through within several days after an oversold bounce is as dangerous as support giving.It is looking more and more like we retest at least $142.50, possibly even the $150-$155 mark in which case I would love to start getting bearish in the GLD again as I still think it can get to $1200 or lower before we flush out all the bulls.

I am watching oil carefully, I was considering buying USO calls and an energy related name today buy I was waiting for a pullback in USO to 32.50. 90 in the /CL works too. I want to get this out before market closes so ending the post now I will add some comments later.

I want names that aren’t highly correlated to the market here as we are stretched, consolidating and near longer term resistance. Additionally, as long as last week’s high holds and we don’t close above it, I am in no way ready to be too aggressively long overall. I don’t necessarily need inversely correlated names all the time, (like the UUP call, FXY call, GLD puts that I have done in the past) but I am trying to find the right balance that keeps me correlated closer to zero for now while still looking at plays that I expect to gain, while still allowing me to profit from a break either way or none at all. LOL, easier said than done.

————-

I got on the AFFY train today as it romped above $1.40 , taking out it’s previous high.  I’ve been mostly following Option Addict when I can on the long side and I passed on AMZN but got GOOG and another play I will keep quiet about… I don’t think OA has mentioned to the IBC crowd just yet today.

I was tempted to grab some gold miners, it wasn’t until late yesterday that I started to suspect they might for once actually not lag GLD and SLV and perhaps should GLD run, could even lead for awhile. But I already have some miner exposures on some metal plays, and FCX which has some GLD exposure (mostly copper though), and I was just a hair late to the punch this morning and they ran away from me.

A lot of the same GLD miners are showing up on the same screens they were yesterday. But I also have REE,ESI,LEDS and some home builders on my radar tomorrow just as sort of a preliminary list.  I am also thinking about adding FRO down here to piggy back off the FLY. I am content to sit on my hands as well, but I should get a better clue as to what the correct play is here soon. If I still foresee more confusion as to direction, I will start looking at some plays in ETFs for longer term that have low correlation with the SPY. DBA on a monthly chart looks like a good spot to bottom pick with the new lows as the stop. Oil on a monthly chart is still in the process of setting up here, but I would say for sure into 2014 it looks good, if not into the upcoming hurricane season. This also may mean good things for the fertilizers and potash names. Here’s a correlation chart of some ETFs to look at for reducing your portfolio’s correlation to the S&P’s movements.

correlation

Or perhaps you prefer a list of ALL names and their correlations for the SPY.

I will have more on the value of “reducing correlations” and such later…. For now just understand that timing is still the most relevant factor, but adding low correlation or negative correlation plays can allow you to stay aggressive without having the same degree of directional market risk.

Comments »

Update

A little late on this update since I’ve been intently watching aapl 445 weeklys putting forth the sideways screw, but I closed remaining weekly SPY put options at 0.25 as they were expiring today had opportunity to get out at about twice that but volume pocket below so I waited and sold at the higher low on 1m chart.

Recap: Bought at .05 sold half at .35, 1/3rd of total position at .60, and 1/3rd at .25.

Still holding last half of the June07 from $0.75 after selling half at 1.61.

Still own DECEMBER SPY puts @ $154 strike from $4.30. But this one differs in that it is more to preserve my capital and allow me to press my longs (that I expect to outperform the market) a bit harder in the meantime if the market turns and I’m busy taking off individual positions I get bearish more quickly. I plan on selling with plenty of time left to prevent theta decay.

Was right about market finding a temporary bottom, wrong about it lifting AAPL. Will likely scratch out for a loss or break even soon.

At some point I gotta get back to picking individual names to be bearish on rather than these SPY hedges. It does work just fine as long as the individual stock and option plays outperform the market in up and down markets, but the alternative of finding individual bearish names that break down to a greater than the market and can outperform in both conditions is much more ideal if you can get it. Maybe LNKD below $170.

I tried one bearish name early today in ASTX. Not perfect but should get the job done more than the SPY should we roll over, and maybe even if we go sideways.

Was tempted to take a bit of my JUNE FXY off today and put it back into UUP while keeping the longer term September FXY on. But the yen is still so oversold and showing early signs of breaking out of a wedge at least on a daily time frame. The Euro and Dollar have yet to clearly identify a direction. The euro has two potential outcomes. One is the upwards trend channel, the other is a head and shoulders breakdown. The dollar looks more like a breakout then not to me, but UUP still needs to clear $23. Both are consolidated pretty tight and ready to move and dollar is hinting slightly at higher and euro lower, but it isn’t clear yet.

GLD needs to close above $135 in the next 2 trading days or I am out of it as well as FCX. It also must not break below the lows before then or I am out.

That’s all I have to say for now. If I get a chance I will talk a bit more about correlations and managing portfolio volatility for greater long term growth.

5/28: When Dow was up near the highs again, this time I grabbed some TRV and LNKD puts. I will be getting rid of a lot of puts if we close this week at new all time highs. Good spot to manage risk to concerned about day-to-day run above new highs. On longer term basis anything in the 15500-16000 level in the dow looks like strength to sell into. I expect a pullback at some point. If this is a secular run starting I will have plenty of time to get back in above the long term megaphon/broadening pattern.

Comments »

FEAR – False Evidence Appearing Real

There are 3 appropriate acronyms for the word FEAR. F.E.A.R.
Fail Early And Often (good advice for gaining experience and learning quickly, and learning how to manage it… but not a good “mantra” to drill into your subconscious mind.)
Then there is the legitimate kind. (F&* Everything…. And RUN!)

And finally “False Evidence Appearing Real”…

The reason I prepare with the charts and such is so I have a good enough read and enough knowledge of the situation to consider MULTIPLE outcomes.

In the most recent case, I was fully prepared to begin shorting at dow 15500-16000 and started actively hedging via short GLD before the first drop, long the dollar (UUP) and long the Yen (FXY). But once 15500 hit which happened to line up with the euphoria in the news that Bernanke would keep QE going, I jumped into action and grabbed some weekly 164 puts. This is what I was looking at this Monday.
dia monday

And I had previously posted this longer term chart…

$indu

Also, “Luv2Gambool” had a great post on the market topping that provided me with a bit more conviction.

I had began adding puts to attempt to call a top from last Friday. Yesterday was my first attempt at weekly options which was an aggressive play. I was surprised to get the fill at 0.05 for 164 puts on the SPY. I put in the order then I brought up the charts and before I could even see where we were, the puts filled. I was getting ready to edit the prior order to .10 to make sure they go through because I saw how the 15500+ range was a great spot to get short and I wanted in immediately. It wouldn’t go through because the trade executed. So I went to put in another order only to find the bid was already above .10. I ended up not putting in a second order on the weeklys, but it’s okay because I sold half at .35 and today sold half of the remaining half at 0.60 as well as half of the puts the first week of JUNE. I then added an AAPL call. WHY suddenly go from expecting a 10% correction to making a bullish in a leveraged name with a leveraged CALL option?

Sentiment. F.E.A.R. It was a lot of “FALSE EVIDENCE APPEARING REAL” Nikkei down 7% reminded me a lot of the Lehman bankruptcy that caused capitulation bottom. Everyone was suddenly so fearful. I also have to tell you, with options, much less weekly options, you can go from neutral to super bearish in your portfolio very quickly when you have a gap lower like this. And I wanted to at least lean closer to neutral after a move downwards… If not bullish. But that was after long periods of a 2 year wreckage this is near the highs. I don’t know, maybe Japan had their “1987” type event. Interestingly enough, it was the 1987 event and talks of manipulating the dollar 50% lower that shifted capital TO Japan leading up to 1989 Nikkei bubble in the first place. Well now there is a “1987 event” 1000+ point down day in Asia…. Does that mean capital could shift back into the US and form a bubble? Of course. Regardless, there is a lot of domestic and foreign capital that got out of the Nikkei and Asia names. There is a lot of capital moving around. That means the potential for big moves. For the time being there still is the possibility that we retest, breakdown, and correct. Or we could have already bottomed. I would keep a close eye on yesterday’s LOW and more importantly yesterday’s CLOSE. It is at this point that I may add some puts with more time on them since we could very well chop around here if this is the worst of the flushout for the next few days. I see a tight range followed by a break. My bias on a more reasonable time frame that isn’t so fickle about shifting around every couple days is short until we break and close above yesterday’s high. My longs have been done with a HALF position size.

Note:I also added GLD calls of all names! 130.75 is where I give up on it. If the lows give GOLD sees a quick drop to 1200. But we are oversold here so I think it will possibly retest the breakdown point before that happens.
I also added FCX calls. Both July dated. I was going to wait until GLD hit 1200 before I added FCX, but I like GLD’s spot. That big pivot move off it’s low makes me bullish as long as we clear GLD 135 in a few trading days (and stay above the low) I will stay in the name. The stop is just below the low until we start moving above 135. Then I raise the stop based on it’s 3 day low. Eventually I loosen the stop if the move is working. Ultimately for now I am looking for at least 140 to the upside.
Also added the VIX exposure (VXX July Calls) as we retested yesterday’s low and successfully filled the gap. Nice inverse head and shoudlers setting up in the VIX too. Lots of moves today

Comments »

Computer Generated Names

I get “trade alerts” via email when certain patterns pop up. I have the alerts filtered to only give me stocks expected to make a 30% move or more.
What I do from there is look up the pattern length in time to give a projection of how LONG it takes to reach a target, and I calculate the return to the bottom end of the targeted range. THen I divide the return by the number of days to get a “return per day” average expected. Then I sort the names. I separate those optionable from those that aren’t and those that are OTC.
Yesterday the list was particularly stunning as there was multiple expected returns over 50%.
The results are below by category in order by return per day expected based on TODAY’s quotes at around an hour after the open

OTC: NEPH,MSTG

Optionable: FSLR, ROSG,MMYT,AERL,VICL,SWS,KLIC,GHM
NON-Optionable: KIPS,LIVE,BIOF,SPIR,MXC,RNN,KWT

Also significant is that KWT a solar ETF is signaling a big move for solar has yet to occur.

Bull Flag patterns and Bull Pennants always dominate the best return per day even if they have lower expected returns because the pattern duration is typically only a few days.

So TU looks good and LINTA looks even better.


I doubled down on my hedge and grabbed some weekly 164 puts for a nickle 0.05 into the euphoria.

3:27pm est edit:Sold half at 0.35

Comments »

Just Some Winship Right Here

Just showing everyone I know a thing or two.

Weekend Ideas Should Be Compiled Here

hattery says:

I’ve overachieved in an overextended market during a historically not so bullish time, overbought RSI on weekly and daily extended into upper Bollinger band and overbought slow stochastics near the upper trend channel and resistance coming up on longer term pattern.
The strongest bull markets don’t give you a chance to buy much of a dip very often, but this is a little bit extreme. Protection is my play… But if I had to initiate some new positions.
Gear up for war with ERJ, KAMN, RGR, maybe TASR and the civil disorder with SWHC.

If you want the high debt equity, low quick ratio BK plays with high short float and low float…

CSTR,ALJ,CSUN,ROYL,BIOF and why not SOL too with solar going apesht.
Low float momentum names with EPS growth of 25+% MELI,VOXX, BBG,SWC,ZAZA

And ACTV and FSLR into the volume pocket
But mostly UUP and a small position in puts on Monday.

———————

BOOM! Take a look at those plays. Just taking a look at some of the gains from Friday’s close. Th

ERJ around 2.5%

KAMN 1.3%

RGR has yet to leave the station. TASR only one down 1% or so but I did say “maybe” by it. SWHC so far is break even.

CSTR up 3.2%

ALJ up about 6%

CSUN up near 90% from Friday’s close to today’s peak

ROYL up near 8%

BIOF up 15% today

SOL up around 35%

BBG up over 10% (although option addict had that name pegged like 10% earlier than me)

SWC nice 5%

ZAZA near 6%

Just in case anyone listened.. you’re welcome!

And in case anyone bought options in those names, you can send me a christmas card. haha

As for the “hedging” it allowed me to hang onto names like BIDU and FSLR and other plays a bit longer, rather than selling.

If you have acquired enough skills you should be able to outperform the market. However, there is always the chance that a liquidation takes just about everything down. Can you produce huge gains on down days? Certainly. But it’s a lot more difficult to account for the rare “flashcrash” type of events and get back to where you would be than it is to make up for a few minor losses because of a few puts you buy. Afterall insurance is cheap right now.

I am still long the dollar and the yen. The “Island bottom” in the Euro is a bit concerning for the time being. I do not feel comfortable aggressively adding to an already reasonably sized position in any currency at this point of time. I will let it come in to add while also keeping an eye on the technical breakout point of $23 on the UUP.

Unfortunately GOLD never worked off it’s oversold position enough for me to be comfortable with shorting it on the way down to it’s retest of the recent low and silver’s break. Those type of plays are great because you can make two profitable bets while hedging slightly as there is a slight positive correlation between gold and the S&P.

So when no real great shorts are setting up except the hard to borrow type or type where option spreads are insane, I am left just playing for alpha, betting that my longs will outperform the market and grabbing some puts on the SPY to preserve my gains without selling and protect myself from a drawdown.

My account still is climbing as I continue to park some of my gains from other plays into mostly longer dated hedges (puts in SPY, calls in UUP,FXY), and some just go back into other plays.

Why pay so much for premium I won’t use? Because if I don’t, I am betting aggressively on my timing and on “not being wrong”. I can do that with individual plays if the setup is right, but not in timing tops and hedging which is a more gradual process.

Timing the top is extremely difficult and I don’t expect to hit this out of the park. Just protect against on possible outcome. (I do have a small position in the more aggressive puts the first week of JUNE as well, but). If it moves against me, I want to protect myself with a large percentage of my position intact. The more time value you have, the more people pay for “potential” even for options far out of the money. So the chance of losing 50% of your premium or 100% are greatly reduced from your short term options, even though the returns will be less as well. Since the idea is to “protect capital” so I can focus on “making and earning capital” with the rest, hedges are generally played more as cautionary plays.

 

Comments »

hedge

Grabbing some spy puts at close as I am overexposed to the long side (or at least more than I would like to be) and don’t really want to sell a lot of anything I am in.

Up a lot from BIDU and fslr today thanks to optionaddict which gave me more of a need to raise cash. Opted for spy puts instead.

Comments »

A Lone Samurai Is Ready To Guillotine Yen Shorts

The Yen has been grinding lower, I was and still am not entirely sure if we bottom here or if we get one final major downward flush out move. But the signs are pointing to a bottom, and it would be hard to see a major flush out when we are this oversold on all 3 timeframes.

Daily
yendaily

Weekly
yenweekly

Monthly
yenmonthly

Falling Wedge Pattern.
Triple Oversold (Got it from “OA” Ask Option Addict if you want the indicator).
At Support of Wedge Pattern.
Near Support of Volume Profile (market of plenty of buyers and sellers below can act as a range of prices to hold it from further declines.)
Volume is Thinner above (A bounce can really move higher quickly).
There are a lot of things going for it. Great spot to manage your risk as well as tremendous upside.

For now I am long FXY calls with my capital from my reduced stake in the dollar.

Comments »