Over my years of trading, I became convinced that the standard black box trading indicators and tools that are part of every generic platform are not only worthless, but are something much worse.
With the Rise of the Machines in the trading arena, algorithms were/are being programmed to hunt and run stops. "Running stops" is finding areas on the Price chart where weak, Levered positions are sitting, and eagerly waiting to book their small loss on any adverse move. The 2 most frequent methods of stop running are:
A) Wait until the wee hours of the morning when there is very little liquidity - Then keep attacking a price level with small but relentless orders until you move price to the desired "Max Pain" zone. At certain price points, many traders have their stop loss orders in the market, so once price hits that level, your stop order becomes a market order to close your position at any price. After cleaning out the stops, prices come right back to where it was, however you are no longer in, and you have booked a loss in your ledger.
B) Wait for some widely followed Economic report or other event where jittery traders are ready to jump in any direction on a moment's notice. The second the report is released you slam the market with a wave of large buy or sell orders, looking to induce panic, and make traders pitch their position during a moment of extreme stress.
The logical places to find these weak hands would be near recent highs or lows or near generic trend indicators like the 50 day or 200day moving average. Needless to say, in a game of Poker, I don't want my opponents to be able to see my cards or read my hand, so I determined to create my own set of trend & pivot indicators.
One analyst I had been following created his own Moving Average Line by incorporating other attributes as opposed to the closing price - which is currently the main data point for moving averages and Channel lines. I determined that it made sense to follow his lead so I incorporated various price attributes such as Open, High , Low, Close, and Volatility to create the 3PC Channel pivot points. We then Average the calculations to come up with a 4 Day, 9 Day, and 23 day Channel and we use this to help determine trend, as well as to provide price levels for support and resistance
Test Your Ideas
For any indicator that we use, we want to know if there is a statistical edge. You are constantly going to hear analysts talk about some event, or pet indicator. You will hear phrases like "Price just crossed the 200day moving Average" or "Commercial Traders Sold 20k contracts last week", but what actually does that mean? To know, you'd have to identify every instance this event occured over a specific time period, and then compare it to a baseline sample of price movement to see if you have a statistical edge or not.
4Down1up
So back to our custom Trend Indicator. One thought I had is that I wanted to see what happens when our 9 day (Fast) & 23 day (Slow) Channel are in the same Direction, either down or up, and then price reverse and closes into the Fast Channel Line. Another way to say this is that if the 9 & 23 Channel are pointed down, all 4 lines will be down - On a reversal in price direction, price would close above the lower channel line. We call this our "4down1up" signal. The inverse would be "4up1down".
My Discovery
In our database we calculated the number of times this "4Down1up" event occurred in Gold. Our daily price data for gold goes back to Feb 1975, so as of Jan 2017, the date of this writing, we had 10,523 data points. This indicator triggered 500 times, or approx 5% of the time.
I now wanted to see across those 500 instances, did price rally as we were anticipating 5,10,20,40, and 60 days out after this price reversal signal triggered? When I saw the results, I couldn't believe my eyes. 480 times, out of 500, prices were higher within 20 trading days - A 96% win rate. The mean gain was $11, and the 70th Percentile gain was $5.80. The results became even more impressive when looking at 2006-2016 data - a 97% win rate, with a 70th percentile gain of $22. So this told me that when this signal hits, 70% of the time, gold will move at least $22 higher as of that days close.
I sat there in amazement - The feeling that came over me must of been something similar to what a Alchymist felt upon discovering the philosopher's stone. I am gonna be rich... Filthy Rich... Stinking, F'n Filthy Rich!
A buddy of mine at "The Bank" (name withheld to protect the innocent - or is it the guilty?) is a quant and one of the smartest guys I know. I needed someone to look over my calcs to confirm I wasn't making some type of moronic spreadsheet error, which would render my amazing discovery useless. I sat anxiously as he reviewed the calcs and data, and he looked at me & confirmed that everything looked good - it's legit. YYYeeessss!!!
"Now all you need to do is the inverse and see what drawdowns were during those same periods - Then you know your potential downside, as well as upside."
Of course... I was so busy thinking about how much money I would be making that I didn't calculate what the risk was relative to my gains. I quickly ran the numbers, and when I saw the result it was a kick in the gut.
Not only were prices higher 96% of the time with this signal, but they also moved lower 96% of the time, with average max drawdown over 20days of -9.32, and the 70th Percentile loss being -4.23. It was basically no better then a coin flip proposition! I was crushed - no money printing machine for me 🙁
To confirm this finding, I went back and looked at all 10,500 days in the Gold dataset and there it was - Clear as day. 1) 20day Max move from any random close was 23.74. 20day Min move from any random close was -21.30. 70th Percentile gain was 5.80 and loss was -5.30.
The data doesn't lie - On any given day, prices on average move up OR down by basically the same increment. It's a 50/50 bet, No different than betting black or red on the roulette table.
While I was temporarily depressed as my fantasy was ruined, I realized I did make a valuable discovery. Each individual bet is a Coin Flip proposition. No matter what you think you know about anything, each bet is a random outcome. And if this is so, what the fuck had I been doing all those years betting such a large percentage of my chip stack on 1 bet...one flip of the coin?
If someone came up to you and said - "Hey, I'll bet you $20k on this next coin flip" Would you do it? Why was I doing it? Why do others do it? No wonder my equity curve resembled a roller coaster ride with dizzying heights, and stomach turning falls. I was plunging on coin flip propositions. Wealth that is attained that way, will never be retained. - Never forget it.
What you make via a outsized bet, will simply be given back in the same fashion when "Lady Random Walk" decides to take her ample booty the other way.
What you make via a strict and disciplined approach to your betting will not only be kept, but it will form the foundation of greater gains to come. As you grow your chip stack, and guard it with the Iron Grip of sound betting techniques, your minimum bet size will grow, and so will your profits when you win. The power of compounding will exert it's force, and over time you will begin making the money you were always dreaming about.