There are two systems I’ve been trading since the beginning of 2016. That’s fifteen months of true out-of-sample, real live experience. I’m pleased to report the results have been very good, and consistent with their original backtests.
Due to many requests from readers, I’m offering these systems as a subscription “trade signal” service, as well as the underlying details of the systems for a flat fee.
“Mean-Dev SPY”
I’m trading a trio of systems on SPY or its leveraged ETF cousins. The core system is called “Mean-Dev SPY”, and does most of the ‘heavy lifting’ in this trio. Because the drawdowns are less than buy-and-hold by a substantial amount, I use a 3x leveraged ETF called UPRO instead.
With a a leveraged 3x ETF, I got a slightly higher MaxDD during the period of 1/1/2016-4/1/2017. Profit however is almost double! Well worth the extra 2% in risk. Example below shows growth of $100,000, no commissions deducted.
| buy-and-hold SPY | mean-dev SPY (UPRO) | |
| Periods | 313 | 313 |
| NetProfit | $20,542.83 | $28,655.27 |
| CAR | 16.23% | 22.49% |
| MaxDD | 9.03% | 11.13% |
| CARMDD | 1.8 | 2.02 |
| Sharpe | 0.9 | 1.22 |
| Exits | 1 | 23 |
| AvgDays | 312 | 4 |
| Wins | 1 | 15 |
| Losses | 0 | 8 |
| PctWins | 100.00% | 65.22% |
| AvgWin | $20,542.83 | $2,301.67 |
| AvgLoss | $0.00 | $1,055.79 |
| Expectancy | $20,542.83 | $1,245.88 |
| ProfitFactor | 0 | 2.18 |
| WinLen | 312 | 4 |
| LossLen | 0 | 4 |
| TradeLen | 312 | 4 |
| Skipped | 99.68% | 47.73% |
“DUST in Wind”
The other system I trade is a 24-hour trade using the long/short leveraged gold miner ETFS, NUGT and DUST. Even though it trades both ETFs, I call it “DUST in Wind” because of the old rock song by Kansas. (Check the video: I love the guitar player in the ruffled blue tuxedo…) This system has a very low correlation to SPY, which makes it a good fit for diversification and reducing overall account drawdown. Because these ETFs are so volatile, I adjust my position size accordingly.
SPY has been hugely profitable over the last fifteen months, and that’s hard to compete with. If I reduce my position by half, I get about the same profit and drawdown (slightly worse) as SPY buy-and-hold over the same period.
But wait…I can make the same profit as SPY and only be in the market 12% of the time with this system? That means I can use that same account to trade other systems too. My money gets a double workout and returns are that much higher.
If the backtest out-of-sample data is any indication, this system will continue to do well even when SPY does not look like it’s just been shot out of a cannon.
Example below shows growth of $100,000, no commissions deducted.
| buy-and-hold SPY | DUST in Wind | |
| Periods | 319 | 319 |
| NetProfit | $19,994.89 | $37,083.64 |
| CAR | 15.49% | 28.30% |
| MaxDD | 9.03% | 22.47% |
| CARMDD | 1.72 | 1.26 |
| Sharpe | 0.89 | 0.8 |
| Exits | 1 | 40 |
| AvgDays | 318 | 1 |
| Wins | 1 | 26 |
| Losses | 0 | 14 |
| PctWins | 100.00% | 65.00% |
That’s a peek at two of the systems I trade. If you’re interested in getting the same signals, you can find out more about these systems here.
4/20/17 edit: A reader asked if I’d share trade dates. I’ve put together a spreadsheet with all of 2016’s trade dates for each system (link).
Can you post your signal dates and trade dates from the past years?
I’ll share the trades of a subset. Here’s a link to a spreadsheet with all of 2016’s trades.