Trading the Tides, Part 2

Last week I posted a few charts showing equity growth if one were to trade the tide turn dates over the last 3 plus years. If you scroll down to earlier posts, you'll see the two charts, one based on what I have termed a 50% entry/exit and the other labeled nirvana trading where the absolute best entrys and exits are assumed (not possible). The reason for these two charts is to frame what is felt to be the absolute best and possibly worst trading results (although it is entirely possible for a trader to do worse than an average 50% entry/exit).

With the extremes set, we should now take a look at the range in between these two. A little more explanation is in order to understand the following charts a little better. When I refer to a 75% entry, here is what I mean. Lets say that an impending tide turn is imminent (for instance this coming monday) such that the trade called for is to close out a long position and reverse to a short position. Also, lets assume that the snp ES range for monday is a low of 1,060 and a high of 1,073. A 75% exit long and entry short then would be 75% of the difference between the high and low, and since we are looking for a long exit/short entry, we want: (1,073-1,060)x.75 plus 1,060. If I have done the math correctly, this translates to a price of 1,069.75. So this price of 1,069.75 then represents a 75% entry/exit for this day.

Similarly, if monday were instead a tide low date (short going into that day) and the price range were the same, a 75% exit/entry to exit short and enter long would then be: 1,073-((1,073-1,060)x.75) or 1,063.25. Pretty simple concept.

Now, lets look at some equity charts with varying degrees of success with average exits and entrys. Below are 3 charts illustrating the results.

At the top of each chart in the title block the entry/exit assumption is shown. Below is a chart with all the charts combined.

The heavy white line has been drawn to show a 100% return on investment, on in this case, starting with a $25,000 account and have it grow by $25,000 annually. Apparently, to realize this type of account growth, the average entry and exit would need to be in the 65% to 70% range based on backtesting data for nearly a four year span. At some point, I would like to go further in the past with backtesting. Again, this return is based on trading one ES contract and has no consideration for drawdowns, which if you are a trader, you would know how those can affect a trader's ability to stay in a losing trade. More on this later.

Mid Week Update and Nirvana Trading

Above is a chart of where we are on the snp ES as of about 5:35 pm, tuesday 10-6-09, the day after my birthday. :) Clearly the tidal low expected for today actually came in late last week. There seems to be a tendency for a turn date to occur one or two days early in the direction of the predominant trend. In other words, in this bull market, the low and turn upward seems to come in earlier than expected, as if the market can hardly wait to resume it's upward march in a bull market. I will have to check to see if this the opposite occurs in a bear market......In any event, monday the 12th is the next high point of this rally according to the tides.

.

Now, back to trading. Last weekend I posted a chart based on trading the tides with a 50% entry and exit. Just for giggles, below is a chart of nirvana trading where the basis is that a trader would get the absolute best prices available on the turn dates. Obviously this is not possible, but I post the chart just to frame the boundaries between what should be bad trading (the previous 50% chart) and what is impossible to do (the 100% chart). Now the limits are set and this weekend I'll put up some charts of equity growth based on what we might theoretically expect a trader, only slightly smarter and more experienced than that caveman mentioned earlier, ought to be able to do. Keep in mind this is based on trading a single ES contract and getting the impossible best prices for entry and exits. Also no consideration for drawdowns is included here.

Trading the Tides ?

Since beginning this blog a little over a month ago, I've touched upon the tides and trends but have said nothing about trading. Let's delve into that area a little bit here. It seems obvious that if the tides are anywhere near consistent pointing to top and bottoms, we ought to be able to trade with this information. So here's what I 've done.
.
My tide data goes back a little over 3-1/2 years to the beginning of 2006. I've downloaded the S&P emini daily price data (open, high, low, close) from Tradestation over that time period and laid it up against the tide turn dates I've developed. Since the eminis trade 24 hours a day, understand that the official beginning of a new day is at 4:15 pm est and runs to 4:15 pm the following day.
.
An assumption must be made in order to backtest the data. That assumption being at what price on each turn day one would theoretically be able to either buy or sell the emini contract. For this assumption, I've varied the efficiency of the entry and exit points to examine various equity curves based on these entry/exit points. I start first with what I call a 50% entry/exit. What that means is that on any given tide turn date, a trader would only be able to enter or exit a trade halfway (50%) between the high and low of that particular day. In other words, if the tides indicated a turn next tuesday (which they do), and the trading range for eminis on that day was from 1010 to 1030 and the trade required was to go long, then the trader's best low entry price for that long entry on that day would be 1020, or halfway (50%) between the high and the low of that day. As that long trend and trade played out to the next turn date, his exit long and go short price would again be 50% between the high and the low of the turn date range. So, if the trading range on the reversal turn date is between 1040 and 1052, then the exit long and go short price would be 1046.
.
Using an excel spreadsheet to tabulate all the entrys and exits for each turn date, I then calculated an equity curve using a theoretical $5 per trade commission cost. Below is a chart of those results dating back to the beginning of 2006 with an assumed starting account of $25,000.


As you can see, nothing special here. Over almost a four year period, a gain of only about $10,000 on an initial investment of $25,000, or roughly 10% annual return. And some fairly wild swings in equity with undoubtedly large drawdowns along the way. Surely, we can do better than this. Well, we can do much better than this...actually much better by improving the entrys and exits on the turn dates. Keep in mind the equity curve above is based on only being able to get average entrys and exits at prices halfway between the highs and lows of each day. Surely, even a caveman could do better. Mid week, I'll put up some more equity curves based on what a trader should be able to do.