To trial or not to trial...

I was considering starting to play with the Williams %R charts with some live money, as is my way, then decided that I had best try something different.

While I know that trading with "virtual" money is not the same, I feel that I know the execution and pitfalls associated with using real money that I should be able to easily duplicate the same effect. Unlike day trading where the pulling of the trigger is so subjective and based on live data feeds these tests are based on next day opening numbers. I don't have to sit in front of the computer to see what I would have got as I either place a market order and get my fill or I place a limit order and see if I get filled. Checking the EOD data or watching the data live makes no appreciable difference as I am deciding ahead of time what the strategy is going to be.

I set these trades up in a free portfolio page on a free data site to track. I'll keep adding to it and see where it goes. The thing about the number of possible trades is that I may end up needing the higher quantity to keep both diversification and win/loss rate manageable. Trying to cherry pick the stocks can very easily backfire.

If I were trying to use options I think my results would be different but I need the stock results to prove the theory then I can apply it to the options afterwards.

I ran a scan for stocks between $10 and $50 with volume over $1M shares per day and have mid 500's to choose from. Running a separate scan on just those stocks looking for %R on the 6 and 18 day period resulted in a dozen or so, most of which were short or multiple short ETFs. I ended up with only four that closed yesterday at the right point and have decent "at a glance" chart characteristics. One was a stock that I bought the option for on Wednesday past, interesting.

So, I can only trade so many at once, and if I traded these stocks I could only do two, so how to choose and what happens tomorrow when I may have four more to choose from. Obviously I have to pare these down from the original 500 to a few that follow the %R pattern well which should leave me with a handful that I can rotate through at any given time.

This is the benefit of the option service that I am currently playing with... but that is another topic, just the fact that I am playing with options is enough as a $50 stock may only cost $1 to $3 for the option, depending upon strike and volatility at the time.

Jeff.

XBI %R chart essentials

Last post I noted the trades for XBI based on the Williams %R. While they worked out very well there were other things going on in the chart worth noting:



Working along from left to right the previous high of near $70 is just off the chart in August of 2008 and the 100 DMA breaks into a downtrend on October 1st. The price is not really trending but tending down a bit. This is where the %R shines while the price is not going anywhere in particular in the medium term.

There is a triple bottom formed between November 2008 and May 2009. The price starts it's uptrend here after and the 100 DMA breaks up in July triple confirming the new trend. This may have broken now as well except that the price has bounced off of the 200SMA which seems to be acting as decent support.

The red circled area is odd as this is where the %R breaks down. The multiple spikes through the overbought line with no substantial drop toward the oversold AND no confirming price move. Coming off of a likely triple bottom I would be hard pressed to take the short trades here so I would tend toward the long side BUT would wait until the first long signal. That signal showed up and preceded the steep move up.

There was a short indicated in September that I missed noting but see now. Seeing as the previous August high did not surpass the July high a short would be in order here. Even if stopped at $2 up (which would likely be the case for a short) I would not have been stopped out. Entry at $55.15, peak at $56.92 then the drop to the exit indicator at $51.20 for the reversal to the current long position.

Slowly I am coming up with a set of rules to go with the %R style trades.

I know many people go on about cherry picking stocks that fit certain indicators but I find that it is about making the indicator fit the stock or ETF price trends. I've done a lot of "mechanical" testing in the past and it always looks good on paper but the execution was lacking. A lot of that was the subjectivity factor.

Now that I am Trading the US markets and I can place proper stop orders and execute odd lot limit and other orders I think that I will give this one a shot on a small scale. Given the average trade profit of $3.59 with this one ETF I could easily trade 25 shares ($1500 peak cost) as I only need a 40 cent move to break even from commissions. I may give this a try with stock instead of options first. Prove the theory and track corresponding option trades to prove that side of the theory as well.

I will pick a slightly cheaper ETF to try first I think, might as well keep the cost to an absolute minimum.

Jeff.

Williams %R application to ETFs

I considered running this on a few stocks to see what would happen but decided to try it out on some ETFs first. I feel that the ETF better reflect the sector and are less prone to gaps and SEVERE adverse moves based on news or a single company's "issues".

So here is a chart of XBI - SPDR S&P Biotech Index for the past year.

The %R indicators are the 6 and 18 day period. The chart is setup for 1600 X 660 which was good for me to work with and shows up nice when expanded.

Here is a rough table of the trades as noted on the chart.


DATE IN Price DATE OUT Price Profit Target
2008-10-28 $47.30 L 2008-11-04 $55.90 $8.60 $3.00
2008-11-04 $55.90 S 2008-11-21 $46.15 $9.75 $3.00
2008-11-21 $46.15 L 2008-12-17 $52.60 $6.45 $3.00
2008-12-17 $52.60 S 2009-01-15 $51.00 $1.60 $1.60
2009-01-15 $51.00 L 2009-02-04 $54.90 $3.90 $3.00
2009-02-04 $54.90 S 2009-02-20 $51.40 $3.50 $3.00
2009-02-20 $51.40 L 2009-02-27 $48.00 -$3.40 -$3.40
2009-03-27 $50.00 S 2009-04-08 $45.20 $4.80 $3.00
2009-04-08 $45.20 L 2009-05-11 $46.55 $1.35 $1.35
2009-05-11 $46.55 S 2009-05-26 $45.00 $1.55 $1.55
2009-05-26 $45.00 L 2009-06-03 $48.00 $3.00 $3.00
2009-06-03 $48.00 S 2009-07-08 $48.00 $0.00 $0.00
2009-07-08 $48.00 L 2009-08-04 $54.00 $6.00 $3.00
2009-08-19 $51.80 L 2009-08-28 $55.00 $3.20 $3.00

Total per share profit was $50.30 taking the full move and without placing any stops other than to secure some profits on the last few positions. All trades were placed based on the %R indicators and the trades were mostly exited, and just reversed, also based on the indicator. also all trades were executed on the next morning's opening price for sake of simplicity. Using a real time chart to time entries and exits could easily be done in order to secure a better fill price, but this just shows that is not really necessary.

The last column is based on closing every trade based on a $3 per share profit target and still gains $28.10 per share. With the reversal trades this would be silly to do as it gives away lots of profits.

I need to investigate the use of options for these same instruments now as a means of eliminating the need to be concerned with stops and enabling me to hold the short and long (puts and calls) positions while transitioning between reversal trades.

Jeff.

A little cleaning house

I reviewed my positions over the weekend and decided to start cleaning some of the trades out that are not in synch with my current plan.

I am targeting trades that are not in my TFSA primarily and trades that are not specifically recommendations.

I decided to play the trades as recommended by the services that I am trying and not to play trades that involve only placing half of the strategy they are using. A number of the trades involve selling naked options to fund the underlying long option purchase for the actual price move. Nice idea but naked options are very risky if not covered by a stock or long corresponding option. So buying just the regular option puts me at a disadvantage as I am not taking advantage of the credit from the premium sold to fund the purchase of the long option. If they have suggested a net debit trade combo then I will consider it as there is open risk at the outset...other than the naked option that is.

I also will stop buying longer term options to try to take advantage of cheap time premium in order to have less EV degradation as the expiry approaches...some trades are the inside month. While the time premium may be cheap, it drives the price of the trade up and therefore the risk is greater and it also reduces the position size that I can take which will drive total profits down. So I will stick with the recommendations.

I see that they have been closing trades and all are in profit, which is nice to note.

I did a check on one service, the active one that I may keep, that has a track record and find that, over all, they make a healthy profit. Picking trades here and there could reduce the profits should I pick the wrong trades, so I will try not to attempt cherry picking.

Today I closed a trade for a loss, about 13% and closed another for a 100% win. Both were outside of my parameters and the winner I should have closed yesterday when it was at 140%.

I adjusted my stats accordingly.

In order to save a little calculation time I setup a daily return formula on my spreadsheet that uses my dates to calculate the numbers. The average is a straight average of all the trades which does not reflect a return on capital, only a return on the capital used in the trade. I would like to see that number up to 100%, which I think is possible. I will put an average trade capital in some time and that number may mean more when relative to this average.

I did some more work on the Williams %R strategy, mostly thinking, and came up with a method that may work quite well. The trouble that I did notice is that it seems to not work well in a transitioning stock. It is great for established trends, up and down as well has horizontals, just not as the stock switches from one to the other. I think that tracking the performance regularly and placing trades frequently will give a good result overall. I am going to check this against a sector rotation scheme using sector ETFs and come up with a balancing plan to weight the trade sizes according to the performance results. I am going to guess that the horizontal pattern will be the best as it allows taking both side of the trade while up and down trends are best only trading with the trend.

Jeff.

Jeff.

Combining a few strategies...Part I

OK, this chart is a bit more complicated than I intended and the final post got longer as well, bear with me on this one and I will break it into two parts.

My last post had an example of using Williams %R on an up trending stock to play the long side. Buying and selling the stock or the call options could be interchanged easily enough.

This time the stock is BK - Bank Of New York Mellon Corp

I added a mean reversion plot (red, slight up trending), 20 day price channels (orange dashed). The blue arrow indicate trade and direction, the blue vertical lines are the indicator triggers and the blue trend lines are just to mark the general price move between trades. I did not number the trades but they are counted left to right sequentially for reference.



This stock is going nowhere in a hurry so I would play both directions on this one. Using the Williams %R 5 and 14 day oscillator as an over bought and oversold and placing trades the day following both crossing the -80 (oversold) and -20 (over bought). Instead of just closing the trade I would just reverse the trade at the same time. There may be some slippage with this plan but if it is run consistent profits are still regular enough to not really have to worry about it.

If trading the stock, just buy at the bottom and sell at the top then immediately short, cover and buy at the bottom. I think there is an easy way to do this... on the first trade on the chart just buy the 100 shares, market order is fine, so filled at perhaps $25.50. On May 6th the indicator is overbought on both periods, place a short the next morning for 200 shares which should close the initial trade and open a new short for 100 shares. This saves on one commission charge as it closes the first trade and opens another in one trade. I have a feeling that this will not be allowed in Questrade as it is specifically a "short" order...technical issue but it garners them additional commissions.

Next trade on May 21 would be to buy 200 shares covering 100 shorts and opening 100 long. Again, one commission.

Trading every single indicated trade on this stock over this period would result in 14 trades, one still being active. The gross profit for the 5 month period would be $33 (rounding down in most cases for easier math). That's a little more than doubling my money on the trade and $3300 for a bunch of 100 share trades. Maximum capital in play at any one time was the triple trade in June for $8850 or so.

Playing the options for the same stock would be about the same method, although I cannot think of an easy way to reduce commissions as long is buying a call and short is buying a put. They can be executed at the same time but I cannot open a new trade while closing another in one step... not without selling naked options...but I won't go there.

So buy the call to play long and buy the put to play the short game. No need for stops and no need to buy long term expiry's as most of the moves have been a month or less...so maybe always buy two or three months out.

Unlike the stock prices these option prices are a little sketchy as they are not tracked as well, nor do they necessarily trade every day so actual tracking of value is difficult at best.

BK calls and puts right now are around $1 and the spread is 5 or 10 cents. Assume that all things remain equal, and due to the lack of trending they have been mostly. Buying a call $1.50 OTM costs $1 (one month to expiry). The price moves $3 (our "target" even though the price swing has been averaging a bit higher than that) so the Intrinsic Value goes up $1.50 and the Extrinsic Value (the purchase price) holds it own (after all the price is still $1.50 away from the strike). Net increase is the $1.50. As the option gets closer to expiry I might expect to lose most of the EV so "holding it's own" is not the case with regards to time flow.

If I go one month farther out the cost of the options only goes up 40 cents but the degradation in EV is slower at this point... I might only lose the 40 cents instead of much more due to the non-linear drop in EV in the last month.

Making some broad assumptions based on current option prices, $1.40 for a two month option that is $1.50 OTM and adding the price move that puts it ITM and deducting 40 cents of the premium paid due to time value decreasing yields a total gross profit of $13.00. Puts and calls seem to be priced about the same due to the lack of a trend of the stock price.

That is $1300 compared to the stock trading $3650 for the equivalent of 100 share trades...when considering the ROI it is a different matter. The stock trades were, on average $2800 per trade whereas the option trades were $140 on average. That is 130% compared to 928% ROI. With the stocks I would not be able to change my position sizing along the way, or just once near the end of the period if I chose. With the options I could, if I wanted to risk only gains, increase my contract size by one contract a few times along the way in addition to using the unused capital for other trades.

Compounding my option trades while not risking anything beyond the first $140 would result in a different outcome altogether as the very first trade more than doubled my money (I already have trades doubling my money so it is not unusual in options).

The first trade nets $460 on one contract so I tuck away $280 (double my initial investment and continue to work only with profits, but with all the profits used to compound my money). I end with $10,325 or a nice 7475% gain. This is the power of options combined with compounding as they compound much quicker. It is sort of like the question about which would you rather, $1,000,000 now or 1 cent now but doubled every day for a month, which equals $5,368,708 and change.

I would expect that I might just increase my position size by one contract each trade which would see $3823 in profits, still a 2730% profit.

Jeff.

New attempt at oscillation trading

Leave it to me to try something else when what I am doing is working fine. This time I will leave well enough alone and NOT change what I am doing with regards to the option strategy involving using the services that I am currently testing. Also I am still funding for the Optioneer strategy, no change there either.

So, on to my next endeavour.

I don't know what to call this but it involves trend trading, sort of seeing trading with a slight twist.

I am plotting a few indicators and overlays that I have been comfortable with for the last while. On the chart below I left the legends on so just click on it for a larger view.

The solid lines are the typical simple moving averages. The dotted lines are 100 day simple moving average envelopes which I was using a while ago to determine entry and stops for trades. I never really used them but they certainly help visualize stops while the price is in the channel.

The oscillators are both Williams %R which is just an over sold and over bought oscillator similar to Stochastics. I am using the 14 and 5 day for comparing two timeframes without plotting two charts.

The bottom indicator is the Bollinger Band,a volatility indicator. I was going to use a standard deviation but it is identical to this.

Here is a stock that I am currently trading using options. It was a trade based on the service that I just recently cancelled as they were providing one trade per week and I didn't feel that they did much more than watch for chart activity. It is worth noting that I bought long options on September 14th...according to this chart it was the wrong time to go long but I still have some time to go before expiry so I will see where it heads next.

ININ - Interactive Intelligence Inc.

I plotted green and red vertical lines to represent buy and sell indications. As the %R in both the 5 and 14 day period dip to -80 the next day the stock is bought at the opening price, that makes it simple but lower limit orders could be used. Seeing as the stock is trending nicely I could have easily just bought when the price came within a certain range of the 50 SMA, a typical swing trade move. I will plot this over other stocks in varying trends to see how it holds later.

The exit is as important, or even more important, than the entry. Seeing as I am using options I plan on buying options with strikes close to the stock price at the time, estimate the timeframe to determine expiry dates for the price move and not use stops. The total trade cost will not exceed my maximum loss allowance at the time. If I were using stop loss orders instead I would use one of the 100 SMA envelope lines or the 50 sma.

The key is when to take profits and this is where this strategy helps as it takes the subjectivity out of the process.

Initial stops are easy enough, use the SMA that closely corresponds to the Maximum Loss Allowance or an unbroken SMA that forms a trendline and move them with the SMA chosen. The stop will move up slowly. As long as the price does not move too far off the lower trend line (no straight lines here, just the SMAS) there is no need to do anything other than adjust the stop every day or two.

Once the price moves then there are two methods of exit:

1)$3 target (or any price of choice)
a) with stop loss exit
b) with limit order exit
2) %R short term dip

The first is easy enough. Using the entry method I believe that a tighter stop than the profit target can be used. In the ININ example a $1 stop could be used initially and a $3 profit target provides a 3:1 profit to loss ratio, acceptable. Once the price moves in my favour I continue to move the stop loss with the SMA.

Option a) has me move the stop loss up to the $3 target as soon as, or very soon after the stock price passes this price. Exit will be by market order once the price touches the stop. The advantage to this is that the stop can be moved up tight with the price in order to secure greater profits should the price move higher. The disadvantage would be the possibility of the price gapping down through the stop.

Option b) has me placing a limit order and getting the $3 profit, or better if the price is higher at the time of the order placement. The advantage is that once the price is above $3 I am pretty much guaranteed that profit. The disadvantage is if the price continues to move up I miss some additional profits.

Option 2) has the exit order place the day following a dip below -80 on the 5 day %R.

There is a third exit, if the price moves slowly up then the SMA stop keeps moving and either gets hit along the way or just keeps following the trend. The loss slowly reduces along the way and may turn into a small paper profit position.

ININ

Three trades plotted in the last six months profit using:

1)$3 target (or any price of choice)
a) with stop loss exit = $12 (trailing by $1 raised in 50 cent increments)
b) with limit order exit = $9 or slightly better
2) %R short term dip = $9.15

No matter the method there were profits to be had, and that is the important part.

I won't go into the possible option strategies here but suffice it to say that options for this stock could be had for a fraction of the cost of the stock.

One thing that I see mentioned about trading with any type of automatic or mechanical style is that they are prone to curve fitting. I think that the problem with washing any strategy as curve fitted, making the indicators work due to the stock price activity, is that it works historically for a stock and that pattern can be used to identify other stocks that are in similar early stages of the fitted strategy. So why not use the fitted strategy and in turn fit the stock by pattern fitting? I have a successful short list of stocks for my P & F strategy using this idea. Therein lies the key, stock selection to match the strategy employed and not trying to make a strategy that works for a few stocks work for every stock or even ones that I may like to trade.

I think that almost any strategy, if proven for some stocks, can be made work in this way. The second key is in recognizing when the fit is no longer good. This is accomplished by keeping records and tracking performance in order to see when the overall strategy no longer works for a particular stock or group of stocks.

Jeff.