The following market commentary is purely related to the systems I operate. To learn more about these systems, please click here.
A somewhat quiet week again with just one trade, for one win in bonds. Admittedly I'm glad trade volume is down right now. Post BREXIT price action is still a bit volatile, even as volatility as measured by the vix has come in significantly.
Next week begins a two-day FOMC meeting which I do not trade through. So it will be another quiet week.
As for overall price action outside of bonds, markets are somewhat neutral with a slight bullish bias. This week has been somewhat of a pause as the monster move up the prior few weeks is digested. It's healthy price action.
Commentary for the week ending July 8, 2016
Quiet week with the holiday on Monday and then ISM on Wednesday and NFP on Friday. So just one trading day, Tuesday, producing a six tick win in ZB.
The timing of economic data is somewhat of a relief. Knowing the system does not trade through those releases meant a bit of down time this week. Which gives the market post "Brexit" time to settle down and get a sense of where it wants to go and how it wants to get there.
Friday's price action was very bullish in equities though the daily chart is not in any confirmed up trend in my analysis though it has a bullish bias. I'd prefer it remains in an uptrend, and volatility comes in further. That's more advantageous for my bond system to trade.
The next few weeks trade frequency should pick up a bit absent no insane price action and with no major economic data releases.
Commentary for the week ending June 24, 2016
Wow. That was some insane price action overnight. When ZB trades more than one tick or ES trades three or four ticks, you know you are in unchartered waters. Needless to say, volatility was high enough that my systems did not trade on Friday.
I did run a test (back test) to see how the day would have traded absent the volatility logic and ZB would have had a win. Reason being, the volatility was really all overnight, at least in bonds. Once equity cash opened, it was a somewhat subdued day (though I am writing prior to the close). Again speaking about bonds.
I'm hopeful markets quiet down before Tuesday when my ZB system is set to trade again. It will be hard to not interject and just keep the system shutdown for a few more days to let markets return to some more normal volatility.
At least bid ask volume returned. In bonds as of this writing it's more normal than last night when there was no liquidity being provided at all. Thus the big moves.
One thing Friday reminded me about. Thankfully I trade intraday. I would hate to have overnight risk in such an unstable environment.
Commentary for the week ending June 17, 2016
Another quiet week with two trades, for two wins over all. I've stopped trading FOMC days after studying five years of historical data. The Tuesday when the meetings start is a somewhat breakeven day historically. But the Wed. when statements are read is a losing day.
So for now, history says don't trade FOMC days which we had two today and a total of about twelve for the year. I'll continually monitor future FOMC days and backtest to see if the trend continues. For now though it appears the market is too freaked out those days resulting in too much trend for a mean reversion system like mind.
I admittedly have Brexit on my mind, which I wish I didn't. It won't influence my decision to run the system or not. All logic to run is built in. If volatility levels are met, the system will trade. If not, it won't. But I wish I would stop thinking about Brexit. Another annoying market event like so many that have come before.
Commentary for the week ending June 10, 2016
It was a quiet trading week, though profitable with two wins and no losses.
I was nervous running the system on Friday with the sharp selloff in ES, YM, TF pre-market. My volatility rules had not been met to skip trading for the day prior to the system trading.
But just a few minutes into the session, a rally in ZB pushed volatility high enough that my system was forced to not trade. And fortunately that was the case. I always study rules by turning them on and off in backtesting and this volatility rule saved a loss. Not just today but Wednesday as well. Always glad to see when logic is favorable.
Here's to some trade volume coming back next week. It should start to pick up again.
It was a quiet trading week. Other than two ES trades on Monday and one ZB trade on Tuesday the rest of the week was flat. There was not enough volatility for my ZB system to trade.
The following week will be a bit quiet as well. I trade all data releases, even FOMC days, with three exceptions, ISM Services, ISM Manufacturing and NFP. All of which are released next week.
Aside from that, I don't see any crazy abnormalities in the markets. Volatility is essentially in a happy place, not too high nor too low. That can change, but right now, all looks fairly good to me.
It's nice to be back trading after a few weeks of sitting on hands due to my circuit breaker. It was a good week with one win in ES and two wins in ZB on the week. Both ZB trades were fairly long in duration though at almost two hours and one hour respectively. I'm usually out of a ZB trade within 15-20 minutes. But these trades lingered.
The correlation between ES and ZB on a tick by tick basis seems to have diminished which is nice. I don't like when ZB moves inversely with ES for every tick. But the correlation is less. At least on an intraday basis.
This week introduced new logic as well as discussed in the Development Journal recently. I'm also nervous introducing new logic to a live system for all it takes is one little mistake with code and you have problems. But I keep changes minimal and triple or even quadruple check prior to going live. Still it's nerve wracking. But all worked well.
Finally, after nearly a month of no trading due to a circuit breaker tripping, my ZB system is back to trading this week. Whether the system stays trading or is shutdown again I have zero idea. Nor do I care.
Volatility is back into the sweet spot this strategy trades best in. I define this sweet spot with two bounds. The higher bound is when volatility is too high. It simply either trends or is so wide, my stops are triggered before my limits fill.
The lower bound is when the overnight range is too low. What I like to call a pinch point for example if you look at a Bollinger band. When the bands narrow, you know a big move is coming. But which direction is a coin toss.
Having accurate backtests I keep to speak of the importance and this recent period of no trading once again reminds me of its importance. Absent such a historical look back, the decision to shut the system down would have been nearly impossible. And absent being able to shut the system down would increase the capital needed to operate this system due to higher drawdowns.Commentary for the week ending April 22, 2016
The circuit breaker for ZB remains in effect, so the system is shut down. I still run backtests each day and am thankful for the circuit breaker to keep me flat.
In the timeframe I trade, there is simply no mean reversion right now. And the moves in ES are wreaking havoc in ZB as they are inversely correlated for the most part.
I've spent a bit of time the past few weeks trying to find conditionals to help me understand volatility better and incorporate into my systems. I believe I found something but need to study a bit further before making the decision to include into production.
When I finalize this, I will highlight / discuss in my development journal. You may think that a mean reversion strategy, at least one that trades for about 5-15 minutes in duration would want very flat price action but not necessarily. My study has been trying to find a sweet spot of enough movement, but not too much. More to come on that subject.
This week there were five trades total. Three wins in ES, one loss in ES and one loss in ZB. My ES system I currently only run one day a week for a number of reasons. One of the biggest is my backtest sample size is currently limited to about eighteen months. This system requires tick data for accurate backtests and I am limited to what I have available.
Therefore rather than open that system up and trade it blindly, I am limiting it to only one day a week. And I chose a day that happens to be most efficient and a day my ZB system does not trade. So it allows me to "double dip" in capital if you will.
The ZB system went into production last summer but I rushed it a bit. After a few losses and weak performance I found some logic that helped performance a bit. I also adjusted my risk reward by increasing the reward while maintaining the risk. This all got me into September timeframe. And then I was hit with a good drawdown in October into November. Which I thought was over in December.
But it's lingered. To date the drawdown is roughly 22% from the previous equity high. And it is within my Monte Carlo simulations so it's "normal." But it is not fun. During that time I spent a lot of work studying the equity curve and how to apply some measures as an additional circuit breaker. Fortunately I was able to find some useful tools which has now made the system 100% mechanical.
Over the past week the circuit breaker was tripped and the ZB system is offline. It may come back online in a few days or could be a few weeks. Fortunately I have backtests that match live trading 100% so I can "backtest" each day I miss production and build my equity curve. This allows me to know when to turn the system back on and when to keep it offline.
It's very hard to resist the urge to turn it on now. But the following quote I recently heard gives me comfort to wait for confirmation.
"We're not concerned with performance. We're concerned with following the system."
Commentary for the week ending April 1, 2016
Trading remains somewhat subdued with three trades taken this week. Two wins in ES and one loss in ZB. Volatility is coming in so trade activity should start to increase.
Also playing into low trade frequency this week was the calendar. My systems do not trade NFP and ISM data releases. I trade all other data, but those three (NFP, ISM Mfg. and ISM Svcs.) are not traded. And ISM Svcs is released on this Tuesday.
I've been spending a lot of time looking for additional logic to work around volatility but so far nothing has shown promise. I know there is something out there I will eventually find, but for now the systems appear to be doing what they were designed for.
Low trade frequency systems have a lot of benefits including experiencing less trades when a strategy is out of sync, but they are also extremely boring. Drawdowns are lower, but boredom is higher. A lot of indicators point to increased trade volume and win rates in the coming months, but for now, it's hard to say if the weakness from late 2015 is completely over or not. It appears so but only time will tell.
Commentary for the week ending March 25, 2016
Only one trade was taken this week, in ZB for a win. Volatility once again was too high for my ZB and ES systems to take any trades.Commentary for the week ending March 18, 2016
Only two trades were taken this week, both in ES and both wins. Volatility once again was too high for my ZB system to take any trades.
Volatility remains excessively high right now. As a result trade activity remains excessively low. There were only three trades taken with one loss and two wins this week.
I came across the following chart recently which graphically shows just how much volatility has increased the latter part of 2015 and into 2016 (cautionary note though, how have there been 80 such moves in 2016 with about 40 trading days, unless this chart counts multiple moves within a day).
I am very thankful for the volatility logic built into my systems. Absent that simple few lines of code, performance would be somewhat weak right now as moves are just too strong for my mean reversion strategies. So I'd rather have minimal trade activity now versus experiencing a drawdown.
I am not a predictor of markets, but do try to get a sense of when volatility may settle down. Equity markets to me look more bullish than anything. As of this writing the SPX is coming up to the 200MA and previous highs are not that far away. If the 200MA is taken out, and safely retested, combined with seasonal factors, it's highly probably, volatility will come in a bit through mid to late May or at least through April. Time will tell.
Commentary for the week ending March 4, 2016
My systems trade through all data releases, even FOMC days. The only data I don't trade through is ISM and NFP. ISM primarily due to the time of the release and NFP due to the volatility. My historical analysis has shown trading through these reports for the strategies I use sharply reduces performance.
Therefore the way the calendar fell, this was a slow trading week with just one trade taken on Wednesday for a win.
The ES system experienced too much volatility (remember volatility is bi-directional movement, an often misunderstood term) resulting in no trades taken.
Moving into next week it is my hope markets start to settle down a little with lower volatility and more sideways price action. At least my systems prefer that. They can trade with trends but they need cycles. Absent cycles, they do not perform well.
Commentary for the week ending February 26, 2016
As previously discussed, my systems perform best in flat, choppy, sideways markets with low volatility. Trends and high volatility are the enemy of these systems. And the current market is demonstrating both of the latter.
Logic I designed in has been performing well in keeping the systems from trading. Absent this logic for the month of February the system would have had 4 wins and 4 losses for a net loss of 4 ticks on the ZB system for example. Instead the ZB system has remained flat and unchanged.
It's not fun to spectate but sitting and waiting is part of the business. At least my systems are doing the thinking, I'm just doing the sitting.
Commentary for the week ending February 19, 2016
Since the systems are mean reversion, they do not perform well in markets that trend or with excessive volatility. My focus over the past few months has been identifying conditional statements that would define such environments and therefore limit or prevent trading.
The current market has been excessive with both trends and volatility and as a result system activity has been very low. Which is good because when I look at performance of the system absent these filters, performance is very weak. So I would rather have limited to no trade activity during such adverse market conditions.
Remember, adverse is used in the sense of how it relates to these systems. Trending systems would likely be performing well right now. At least those which are bi-directional in approach.