This answers the previous questions about stops. Well, why use stops when you have your magical patterns.
How does a long-only system survive in a bear market?
What happens if the TQQQ kept going down? Do you use stops?
The short answer is yes, I use stop loss. Synthetic TQQQ is comprised of five individual sub-strategies. Each sub-strategy has multiple of its own exit rules. Some sub-strategies have a stop loss rule. However, in a downtrend, the stop loss price rarely gets hit. The reason is that the mean-reverting sub-strategies will take a loss to exit a position during the bounce back before the price reaches the stop loss level. Since ETF TQQQ is composed of 100 Nasdaq stocks that makes it very rare to have multiple consecutive down days without a single bounce back. Backtesting 20 years of QQQ shows that Synthetic TQQQ has had positive returns every year, including the 2000,2001,2002,2008 bear markets. I used QQQ for the 20 years backtest since TQQQ only has about 10 years of data.
In order for a long-only system to survive in a bear market, the key point is to first determine whether it is a bear or bull market. I use a Bollinger Bands based indicator to determine the market direction. Please see the chart below: blue indicates a bull market and red indicates a bear market. Instead of just avoiding trading using long strategy in a bear market, one of the Synthetic TQQQ’s sub-strategies trades when the market is deep in a bear territory, which is shown in the below chart with the red dots. It determines the entry point by combining the Nofry downtrend patterns (see the book, Success in Commodities…The Congestion Phase System” by Eugene Nofri) and mean-reverting strategies. Those trades have a high probability of winning in a bear market. The blue dots in the chart are used for another sub-strategy which only trades in a bull market.
@Tony_Pei, looks like a pretty rough day for your long only system without stops. If we were in a bear market how does your system make a profit out of that type of market?
Notice we haven’t had one for quite a while so bullish strategies will make money but we could be entering a new phase of either sideways or downwards movement. Any statistics on bear markets for your system?
You kill me, Arbi! 
I was going to try to defend this guy a little this morning, despite it being a martingale system. Unless he got out last week, like my strategies did, any stop loss would have triggered with a 600 point loss at the open this morning which wouldn’t be ideal. But then I saw his closed trade and he continued to martingale into his losing long positions this morning.
Its scary @EthosPortfolio, it takes great skill to average down so its not surpising that there is no ‘get out point’ for this system.
But markets are rising overnite as I type this message so who knows he may get lucky and come out with a profit still…but its scary knowing what would happen if the markets had continued to drop.
I was looking for the developer to reply as to how he would have handled a bear market. If there was no reply I would assume there is no provision to handle such an event.
I believe he closed this trade and took the loss. Hopefully he didn’t go on a tilt and go short on a 1000 point down day.
I don’t use a stop loss on my strategy either. I use a reversal in one of my indicators to tell me when to get out of a losing trade. But I also don’t rely entirely on indicators when there is major news headlines affecting the markets. Last week my system was still suggesting that we had a little more room on the upside. I didn’t like how the news was setting up for the weekend so I pulled the plug and went to cash. Too bad though. A couple more up days and my system would have gone short into this pullback instead of hiding in cash. Can’t complain too much though.
I would like to share two messages I sent to subscribers to explain how Senthetic TQQQ works.
Sun, Feb 23, 2020, 11:04 AM
Dear Subscriber,
First, thank you for subscribing to Synthetic TQQQ and for your confidence in this trading system.
Last week the TQQQ dropped 9% from its highest and Synthetic TQQQ suffered about 7% in C2. I completely understand how you feel, since I also trade this system in my own IRA account. I traded 200 shares per sub-strategy which is 200% scaling. Even though the market could always drop further, especially with the recent news concerning the coronavirus, please don’t panic. I will continue to strictly follow this system’s signals. I trust this system, that is why I put my own money in it. My strong confidence in this system is based on the 20 years of backtesting results that support it. During those 20 years, the market has experienced various bear markets. The chart patterns of bear markets are similar regardless of the reason. The system is based on chart patterns and the backtesting results have shown that each year has provided positive returns. However, to prepare for the worst, I will reduce the share size for each sub-strategy from 100 to 80 for new entries. If you are interested in the backtesting results of this system, please provide me with your email address, since this messaging system doesn’t seem to work well when the message size is large. Best Regards, Tony
Mon, Feb 24, 2020 9:58 PM
Dear Subscriber,
Today was a rough day for the market and Synthetic TQQQ. Before the market opened, TQQQ had already dropped 10.96% from last Friday’s close, and Synthetic TQQQ held 300 shares at that time. One sub-strategy had a buy signal at market opening. As I explained in my last message, I followed the signal and bought 80 (instead of 100) shares at the market opening at $95.88. Surprisingly, I got a margin call from C2 and C2 sold all 380 shares at 10:01 A.M. at $99.26. Fortunately, due to the margin call, I gained an extra 380 * (99.26-95.88)=$1,284. After the margin call, I messaged C2 for an explanation, and it turned out that TQQQ requires 75% margin from the long side. Right now TQQQ is $97.89 and this model account has 28K. So this account can hold, at most, 380 shares of TQQQ. That means in order to trade five sub-strategies, each sub-strategy can only trade about 70 shares. So I decided to trade 50 shares for each sub-strategy for now. At noon, I bought back 50 x 4 = 200 shares at price 94.56. Before the market closed, one sub-strategy exited, so at the end of the day, three sub-strategies stayed on the market with 150 shares. Eventually, at closing, Synthetic TQQQ account dropped 6.7% while the TQQQ dropped 11.52%. Best Regards, Tony
No one cares about your stupid explanations. I hope there is no such stupid person who would pay for such a loser strategy
I think most of us were thinking this but didn’t want to say it.
I don’t think there’s any fear of that.
The biggest fallacy with this strategy, besides martingaling, is believing surviving a 2% pullback is somehow “surviving a bear market”. This is the problem with a lot of traders who have never traded through a 20% bear market.
Yup. The last time we had a bear market was so long ago. Most systems don’t even backtest that far! …lol.
For such a system that is always long it will need a good indicator to determine when is the next bounce. Which I doubt there is one is this case.
2 Likes
"Yup. The last time we had a bear market was so long ago. Most systems don’t even backtest that far! "
Very true. As far as I know my system is the only system that has backtesting from 2008 till now and it was profitable in 2008.
This is why I am finishing February in (very small) profit.
Although, to be fair, there’s not many portfolio managers out there that can time their margin calls to be profitable lol
1 Like
Hi @Tony_Pei, what is your comments on this bear market so far with a long only system? Any plans on how to deal with such a crazy market?
Any idea why your strategy’s drawdown is mirroring the market despite the backtests suggesting it would profit from the volatility? I know you put a lot of faith in those backtests. Any chance you could run a backtest on your strategy from Jan 1, 2020 till now? I wonder if the results of the backtest would match the real life performance or if the backtest wasn’t realistic?
