Archive for April, 2009

FAS and FAZ price decay

April 30, 2009


Here is a daily chart plotting FAS and FAZ. I’ve highlighted the price points at which they’ve met recently. I’ve drawn a regression-like line depicting their mutual decline.

The bottom study is a correlation study with a 30-bar lookback. As you can tell, I need to upgrade my image publishing skills. If you can’t read the values of the correlation, they start at -0.75 on the top and go down to -0.95 on the bottom. The current correlation value is -0.81. Not perfectly negatively correlated, but pretty close.

The prices at which they’ve met in chronological order is about $9.52, $9.17, $8.35 and $8.22. They started these shenanigans around April 13, 2009.

One last drag, and then I’ll quit

April 29, 2009

I’m not a cigarette smoker, but I must be smoking something. I announce the suspension of discretionary trading (on a temporary basis), and all I do is discretionary trades. My latest transgression is a quick 3-day trade. I’m in, I’m out. Small profit.

Here’s the play. We’re all gonna die because of yet another natural disaster. This time it a pig flu or something. Whatever. SARS, bird flu, warming, cooling, pandemic over here and polar bears dying over there. I know when I’m being played.

So the latest trade is to buy drug companies (read: the hero of the fable) that will save us from the latest nefarious villain to assault the good people of Planet Earth. Alright, I think I’ve seen this episode before. If not, it looks very familiar. Hmmmm. Anyways, what comes up on the recommended buy screen is NVAX, a drug company, presumably. What do they do? Not sure. And neither do the herd buying this, pardon the expression, pig. A stock stuck in the doldrums of pennydom all of sudden is the hot pick of the day. Well, I see volatility spike to 257% in JUN and the 5 call is selling for $0.75. Okay, I’ll sell that. I’m taking on unlimited risk to make a measly few cents, but I’ve already disclosed I’m smoking something that is probably not good for me.

The trade goes my way on day one, up $0.40. Day two it goes against me as the stock rallies into the close, but still up net $0.20. Day three comes and this flu scare still has legs. The government is talking about spending (big surprise there) some serious money to control this problem.

Now I’m getting nervous. What does NVRX do again? Quick Google search reveals they’re working on a vaccine. Uh-oh. It’s time to bail. Put in a bid to buy back my short call for $0.35. Market moves my way in the morning and I’m happily filled. Net profit of $0.40.

Now I’m back to suspending discretionary trading on a temporary basis. At least until the next time. And I’m not counting the FAS/FAZ anti-pair covered call trade, which I still have on. It’s complicated, okay.

Discretionary trading unsuspended (temporarily)

April 29, 2009

Okay, I know I’m going down the road of system trading and I’ve suspended discretionary trading (on a temporary basis), but sometimes a trade looks like it can’t lose.

At least that’s what I thought with a seemingly brilliant hedged pair: long equal shares of FAS and FAZ. All one needs to make this trade work is to buy them when they’re trading near the same price and pray for one of them to breakout. The basic idea is that if they move in equal percentage moves, the one that moves higher will compound itself while the laggard slows its Zeno-like approach to zero. Instead of the traditional pairs trade where one bets on a reversion to the mean, and for high flyer and low rider to get back together, this one relies on the opposite to happen. A divergence of the underlyings makes this a winner. It’s sort of an anti-pairs trade.

So I bought FAZ at $12.00 and FAS at $6.20. Not exactly near the same price, but why get picky here. Wouldn’t you know it that the two devils decided on a rendezvous at around $8.40. That’s a $3.60 loss on FAZ offset by a $2.20 gain in FAS, for a net loss of $1.40. Okay, it’s a rough start. But this is as bad a loss as one can expect, right? After all, if FAZ continues to drop, FAS will continue to rise and before you know it the trade will be back to break-even.

But someone must have figured out what I was trying to do and decided to let these two languish near each other, which is the absolute worse thing that could happen. And to make matters worse, it appears that both FAS and FAZ have some sort of price decay built in because I seem to recall they met around $9.00 just a couple weeks ago.

What does one do with long stock that isn’t moving. Well it seems the only sensible thing to do is sell call options against it, just outside the money at a place that is affectionately called ‘junk’ by premium sellers. So that’s what I did. I sold the MAY 10 call in FAZ for $1.10 and the MAY 10 call in FAS for $0.61. The grand total is $1.71.

Now the worst thing that can happen is for FAS or FAZ to breakout. Depending on which one did it, I would either lock in a loss of $0.29 (FAZ) or realize a gain of $5.51 (FAS). Mind you that accounts for only half the trade. Where the tanking stock settles determines the trade’s total PnL.

The next time I take this trade is when both FAS and FAZ are trading within dimes of each other. Then buy both and sell OTM calls on both. Basically, you get double the premium on the covered call because only one can go up at a time. You still have the risk of the selloff victim, but does anyone really see either one going to zero?

DIA 70 put calendar closed

April 27, 2009

Originally taken as the MAR/JUN 70 put calendar, I got one roll out of it before I decided to close the trade for a small profit.

The initial debit of the calendar spread was $1.99, and that represents the total risk of the trade. The first roll bought back the short MAR 70 put and concurrently sold the APR 70 put. Credit received was $1.33. At that point, the trade looked promising as 2/3 of the risk was taken off with two potential rolls left in the trade.

But then the market rally of the century took place and ran my short APR 70 put to a nickel, at which point I bought it. That short front-month put is a hedge against time decay for my back-month long 70 put. Once that hedge disappears, you’re left with a naked long put, in this case a 70 put in JUN. Long puts with no hedge are not long-term viable. Not only do you need a move in the direction of your put, but you need a big move. Working against you is time decay, which erodes the options value as days go by.

Once of calendar trade falls apart and you’re left with an orphan long option, you either close the trade or assume the time-decay risk of an OTM option. I closed the trade and sold the long option for $1.32. Adding up the credits, $2.65 received. Adding up the debits, $2.04 paid. The net was $0.59 credit on an investment of $1.99.

Even when calendars do not ‘work out,’ you still can get a 30% return on investment. And that’s not annualized. The trade’s duration was about 12 weeks.

Discretionary trading takes a back seat

April 19, 2009

I am suspending (on a temporary basis) my discretionary trading. It doesn’t take much to gather that I’m not making a lot of money doing it, otherwise I wouldn’t be suspending it.

For the record, I’m not suspending discretionary trading because I blew out an account. But I did see a 20% return dwindle to about 5% for the year. Still profitable, but that’s a hell of a lot of work for 5%.

Discretionary trading is an art form. There are some wildly successful discretionary traders throughout history and no doubt among the living today. In fact, you may be part of this elite group. I’m not. At least not now or in the foreseeable future. So as a result, I hereby announce myself as mostly a system trader. I’ll throw a discretionary trade out there every now and then, but they will be small trades.

My pursuit of system trading will include the following liquid, leveraged trading vehicles: options, futures and currency pairs. They will be strategies that have defined entries, exits, stop-losses and money management or position sizing. There will be a logic behind each system that seeks to take advantage of market behavior. These systems will not express an opinion, as must be the case with discretionary trades, about near-term market direction or lack of direction. Instead, they will express a sort of meta-opinion on how markets work and what patterns they display. The only claim these systems have is positive expectancy over a statistically valid series of trades. In other words, there is no expectation that the next trade will be a loser or a winner, just that after 100 trades, the equity curve will rise.

Discretionary trades are different from system trades in that they do have the expectation that the next trade will be a winner. If it’s not, there are various ways of dealing with it on a psychological level, but overall it’s a disappointment.

Discretionary trades are quite a bit more exciting that systems trades and require the mastery of a much wider range of skills. System trades require statistical research and testing and are relatively boring, unless you consider a rising equity curve exciting.

My research into systems trading is going into three main areas. First on the agenda is Trend Following Breakouts. This is easy to program and backtest because triggers for entry and exit are normally some technical indicator whose mathematical equivalent is simple to code.

Second on the list is Market-Neutral Option Spreads. This includes your plain vanilla Iron Condors and the like, but there are two ways to put them on. First, you can use a rage of criteria such as 28-44 days out and decide if you should even put one on this month. This is discretionary. A system-based approach will always put the trade on month in and out, and at the time with the same risk parameters. That is the only way to backtest, optimize and walk-forward an approach.

Third on my list is a non-reversion Futures Pairs strategy. This one is a bit more esoteric and its triggering mechanism is not easily defined. It sells the dog and rides the pony, and has no expectation that there will be a reversion to the mean.

So basically I have two trend-following concepts and one mean-reversion concept on the board.

More to follow.

SPY 83/85 call spread goes to full loss

April 19, 2009

The APR 83/85 short call spread went to a full loss on expiration Friday. The loss amounted to $1.45 on the 2-point spread because of the initial credit of $0.55 that was received on March 20, 2009.

When this trade was taken, there was 28 days to expiration. At the time, SPY was trading just under $77. It closed on expiration Friday just over $87. That’s over 12% in about a month. I won’t bother annualizing that rate of return. Unfortunately for this trade, that is a rally it didn’t expect. Using the traditional methods of calculating probability of success, there was a 77% chance that SPY would not hit my short strike of 83 in the last 28 days. The chance that it would exceed $87 stood at 9%. Hmmmm. This is enough to make one ponder the validity of probability of success as a statistical measure. Or maybe we’re just experiencing some very volatile times. Let’s investigate.

The VIX, which measures volatility in the S&P 500, would be the place to look for what happened to volatility during our 28 days. Remember, SPY is an ETF for the S&P 500, so VIX is valid measure of its volatility. VIX closed around $46 on March 20 and closed at around $34 on expiration Friday. That’s a decrease of over 26%. Nope, we cannot attribute this unusually large move to an environment of increase volatility.

We can let it go and attribute it to bad luck. After all, we limited the risk on this trade and knew going into it what the loss could amount to. We took a trade that had a high probability of success with an additional edge of positive theta (time decay).

At the heart of calculating the probability of success is a Gaussian distribution of price around a moving target of standard deviation. In the option world, standard deviation is volatility. The theory behind option writing, or selling option spreads, is that the market is efficient at distributing information to all its participants and that price action follows a random walk. In efficient markets, it’s not possible for a trader to gain an edge and actually profit in the long term. So why even pursue selling options? The reason is theta. The edge premium sellers have is time decay. Maintain a positive theta position and you’ll gather profits over the long term.

It’s no great feat to persuasively argue against efficient market theory and against a random walk theory (prices have no memory). I’ll leave that for another time. Right now I’ve got a bone to pick with probabilities based on market volatility. Every option premium seller will acknowledge the Black Swan. Essentially, a statistically improbable event that should almost never happen, but happens all the time in markets. I read that there were 12 greater-than 4 sigma moves last year, including a fair share of 5 sigma and 6 sigma moves. A sigma is a standard deviation. The classic Gaussian bell curve puts the probability of a 3 sigma move at the very tips of the bell curve and ascribes its probability to less than 1%. The quick fix for that in trading is to describe markets as having fat tails. And the way the option world takes care of this pesky problem is to point to all the non-zero bid options way, way out-of-the-money. When trading the options in the middle of the bell curve, you get no credit for fat tails. So whether you use delta of the option or a fancier calculation to determine the probability of success, you need to understand going into the trade that your chances are dependent on the arrival of Black Swans.

A quick note on how this trade was closed.

Since I was short the 83 call, it was assigned to me. In other words, the holder of the option wasn’t stupid enough to not cash-in on a winning ticket. For a brief moment I was short SPY at the price of $83 a share. But that was a brief moment as my long 85 call essentially canceled out the short stock position by being exercised. My broker automatically exercises long in-the-money options so there was nothing I needed to do. Except gaze in horror at the week’s rally.

Risk taken off of APR GLD butterfly

April 3, 2009

Well, it wasn’t a classic butterfly. It was a skip-strike butterfly. Or an unbalanced butterfly. Or if you prefer, a regular butterfly with an embedded short call spread.

The original trade was an APR GLD call butterfly with strikes at 99/100/103. The 100 strike is the short strike and has double the options of the outside long options. We did a 3-lot so the distribution is long 99 (3), short 100 (6) and long 103 (3). The risk of a classic butterfly is only what was paid, which is usually about $0.15 for a 1-point wide spread. Our butterfly gave us a credit of $0.40. To receive that credit, we took on more risk than a classic butterfly.

A quick explanation of the construction of a butterfly. Essentially, it is convergence of a long spread and a short spread with the convergence being the short strike. You’ll remember that you pay for long spreads, and only risk what you paid. You get a credit for a short spread, and the risk is the distance between the strikes minus credit received.

Our GLD butterfly is essentially a long 99/100 call spread and short a 100/103 call spread. It’s a risky butterfly because the potential benefit of the long spread ($1.00) is less than the potential risk of the short spread ($3.00). By assuming this risk, we received a credit of $0.40 for trade.

The idea behind the trade is to take off risk when the opportunity presents itself, and have a free lottery ticket. For that to happen, we need to pay less than $0.40 to neutralize our risk.

There are a few ways to take the risk off.

1. Purchase the 101/103 call spread.

This squeezes the short call spread to the same length as the long call spread. When we buy the 101/103 call spread, we buy the 101 and sell the 103. We don’t have a position at the 101 strike, so that will be a new long strike. We do have a long 103 call, so the spread will cancel out that position. Our resulting position is a 99/100/103 call butterfly.

2. Purchase the 97/99 call spread.

This stretches the long call spread to the same length as the short call spread. It gives us a bigger range to profit as the resulting butterfly would be the 97/100/103.

3. Buy back short options.

If we bought back all the short call options, we would be left with a long 99 call and a long 103 call. GLD settling over 99 is good, over 103 is double good. Buying back short options is normally a good practice, but we can buy back half the options and still accomplish our goal of taking off risk, and keeping a lottery ticket. Let’s think about it this way. We leave alone the short calls associated with our long spread and buy back the calls associated with our short call spread. We take out the guts of the butterfly and are left with a long call spread and an orphaned long call.

For this trade to work out as planned, GLD needs to initially trade lower and give us a chance to take off risk for less than $0.40 per butterfly, and then close above 99 by April expiration.

We chose option number 3 and BOT half our short strike options for $0.15. That gives us a credit of $0.25 for a long call spread (99/100) and an orphaned long call (103). This also generates the least in commissions since we are transacting only one strike and not two, as is the case with spreads.

Who knows what will happen in the next two weeks. In any case, we can move on to the next trade as this one has locked in a small profit and has no risk.