NexusFi: Find Your Edge


Home Menu

 





Momentum Based Algo Swing Trading


Discussion in Trading Journals

Updated
      Top Posters
    1. looks_one ShadowFox with 24 posts (60 thanks)
    2. looks_two Schnook with 2 posts (3 thanks)
    3. looks_3 Sandpaddict with 2 posts (2 thanks)
    4. looks_4 kevinkdog with 1 posts (2 thanks)
      Best Posters
    1. looks_one ShadowFox with 2.5 thanks per post
    2. looks_two kevinkdog with 2 thanks per post
    3. looks_3 Schnook with 1.5 thanks per post
    4. looks_4 Sandpaddict with 1 thanks per post
    1. trending_up 6,829 views
    2. thumb_up 72 thanks given
    3. group 27 followers
    1. forum 33 posts
    2. attach_file 27 attachments




 
Search this Thread

Momentum Based Algo Swing Trading

  #1 (permalink)
 
ShadowFox's Avatar
 ShadowFox 
CO/USA
 
Experience: Intermediate
Platform: TradeStation, Multicharts
Trading: Stocks, Futures
Posts: 129 since Jun 2020
Thanks Given: 70
Thanks Received: 157

I have been Algo trading live for about 5 years. My main strategy over this time period has been swing trading sentiment stocks. I trade a portfolio of stocks with high volatility and swing trade momentum moves. While this strategy has been quite profitable, I am attempting to branch out into algo futures trading for a couple of reasons. First is the margin efficiency of futures over equities, second are the tax benefits, third is the diversification possibilities, fourth is the ability to go long and short, and fifth is the upkeep and management of the system.

My equities portfolio is time consuming to generate because the success of the methodology depends on volatility in the individual stocks. With futures, the leverage provides an additional advantage with normal market moves, meaning I would only need to find markets that moved well for the system over a long term basis a few times a year (or less).

When I started algo development, I thought that I could build a complex system using indicators for entries and exits. Probably where every algo developer starts. I searched for every indicator under the sun that might give me an edge. I used every stock indicator given be Tradestation, and every custom indicator I could find. I filtered, backtested, revised, etc. for months and never found a statistical edge that I believed in. I found many systems that worked for short periods of time, but nothing that worked over a long enough timeframe to be consistently profitable (given normal drawdowns).

It wasn't until I zoomed out and stopped looking at indicators that I found my edge. Go figure, nothing in the stock setup provided by Tradestation gave me a long-term edge . I developed a system using daily bars to predict bullish momentum of the next day. If the criteria is met, I take the trade and stay in for at least the entry day and exit when momentum dries up. No profit targets, the only exits are stops that move on a daily basis or momentum shifts that signal to take profit or cut the trade. Also, the system only trades long for two reasons. Bull market and sentiment stocks tend to run more predictably to the long side, and the second is hard to barrow limitations with a lot of sentiment stocks.

This leads me into my futures development. Initially I wanted to take advantage of the large swings intraday in the NQ for instance. I went back to development and generated many systems that gave me a short term edge, but when I attempted to make these work on larger data sets I could not generate consistent profits. I fell back into the trap of using indicators to attempt to filter, backtest, revise and nothing was good enough for me to trust over a long term basis.

So again, I zoomed out and realized the system I had already developed for equities worked great for futures. It worked on every market I tested it on. It worked both long and short. Plus instead of using full contracts as was originally planned I could trade micro contracts to generate the same margin efficiency and profit that some of those shorter term systems were using.

With backtesting and development completed I am ready to go live with the system. The system has just experienced a nice normal drawdown that I believe will make a great place to turn the system on (next post will touch on this).

As it stands now, the system will trade 1 micro contract of MNQ(NQ), MES(ES), MGC(GC), QM(CL), QN(NG), YC(C), YW(W), and YK(S). This may be adjusted as the system progresses to better diversify. For instance, I believe the system is weighted too heavily in financial markets currently but those are performing the best on the system, so I am ok with the initial weighting of 1 micro per.

I am not writing this to give away the systems intricate details. The point of this journal is to attempt to track my progress with the new system and hold myself accountable to trading it as designed. I have found in my algo development that you do not want to override the system. You may think you are smarter than the computer and sometimes it will work out, but eventually you will miss out on what makes the system a winning system.

Initial capital: $9300
Max drawdown: $1800 (system will be turned off if account balance falls below $7500)

Visit my NexusFi Trade Journal Started this thread Reply With Quote

Can you help answer these questions
from other members on NexusFi?
Deepmoney LLM
Elite Quantitative GenAI/LLM
Exit Strategy
NinjaTrader
Futures True Range Report
The Elite Circle
The space time continuum and the dynamics of a financial …
Emini and Emicro Index
ZombieSqueeze
Platforms and Indicators
 
  #2 (permalink)
 
ShadowFox's Avatar
 ShadowFox 
CO/USA
 
Experience: Intermediate
Platform: TradeStation, Multicharts
Trading: Stocks, Futures
Posts: 129 since Jun 2020
Thanks Given: 70
Thanks Received: 157

The futures system was developed around full size contracts as getting accurate micro data for some assets is difficult or impossible. The current version of the system trades micro's based on full size contract data. So NQ for instance is using NQM21 data to trade the MNQN21 contract currently. It always trades the front month. I will need to navigate rollovers as they come.

My basis for my hard drawdown stop and when to go live with the system comes from a long term backtest since 1/1/2015 and a short term backtest from 1/1/2021. The longer term backtest shows my max intraday drawdown is around $17k for full size contract. Therefore I will use $1800 for the micro version of this system. The shorter term backtest shows I have recently undergone a nice short term drawdown that falls in line with the normal expected drawdown of the system. Therefore I believe this is a good place to start trading the system live.




Visit my NexusFi Trade Journal Started this thread Reply With Quote
  #3 (permalink)
 
Schnook's Avatar
 Schnook 
Munich, Germany
 
Experience: Advanced
Platform: Sierra Chart
Broker: Interactive Brokers
Trading: liquid products
Posts: 570 since Jul 2016
Thanks Given: 1,166
Thanks Received: 1,917


Looks and sounds great - hope it works out as well as, or better than expected!

Out of curiosity, roughly what is your average hold time?

Also, do you have certain VAR limits in place? In other words, if all correlated instruments fire off buy signals at once, do you pile in, or do you manage that risk somehow?

Reply With Quote
Thanked by:
  #4 (permalink)
 
ShadowFox's Avatar
 ShadowFox 
CO/USA
 
Experience: Intermediate
Platform: TradeStation, Multicharts
Trading: Stocks, Futures
Posts: 129 since Jun 2020
Thanks Given: 70
Thanks Received: 157


Schnook View Post
Looks and sounds great - hope it works out as well as, or better than expected!

Out of curiosity, roughly what is your average hold time?

Also, do you have certain VAR limits in place? In other words, if all correlated instruments fire off buy signals at once, do you pile in, or do you manage that risk somehow?

The average hold time for the portfolio is 15 market hours, but most are overnight (or over session break holds). Generally speaking, most positions are opened sometime on session 1 and closed session 2. It does trade the full session not just RTH. Some will go for 4-5 days though if conditions continue to be met. Some positions are obviously stopped day 1.

I do not have any VAR limits currently. Each asset is designed to be its own system in the greater portfolio. No filtering or optimization has been applied to the overall portfolio. The initial idea was that each uncorrelated market would hopefully make up for drawdowns in other assets. Thank you for bringing this up as I still have some thinking to do on the diversification of the system and how I want to manage risk on correlated markets. I don't have an answer at the moment. The goal currently is prove that it works as a portfolio system, then optimize the overall system once I learn more about the individual results.

Visit my NexusFi Trade Journal Started this thread Reply With Quote
Thanked by:
  #5 (permalink)
 
ShadowFox's Avatar
 ShadowFox 
CO/USA
 
Experience: Intermediate
Platform: TradeStation, Multicharts
Trading: Stocks, Futures
Posts: 129 since Jun 2020
Thanks Given: 70
Thanks Received: 157

5/3/2021

Closed Trades: None

Open Trades: $146.25
Long MGC $165
Long QM $37.5
Short MNQ $15
Long MES -$71.25

Visit my NexusFi Trade Journal Started this thread Reply With Quote
Thanked by:
  #6 (permalink)
 
ShadowFox's Avatar
 ShadowFox 
CO/USA
 
Experience: Intermediate
Platform: TradeStation, Multicharts
Trading: Stocks, Futures
Posts: 129 since Jun 2020
Thanks Given: 70
Thanks Received: 157

I did some more digging into the correlation of individual assets with the strategy. To my surprise its way less correlated than I thought. NQ and ES give a decent correlation on a daily basis. But nothing else comes close to having any kind of correlation on a day to day basis. On larger timeframes you start to see more correlation but that just means they are all making money over a larger timeframe so not really much to make of that.

Note: this is not a correlation between the markets themselves. This is only a correlation of my strategy to itself on given markets. This is not saying that Corn isn't correlated to wheat. All it is saying is the system doesn't see much of a correlation between trading corn and wheat given my strategy.

You can see this in today's trades actually. I entered long MES and short MNQ. Go figure


Visit my NexusFi Trade Journal Started this thread Reply With Quote
  #7 (permalink)
 
ShadowFox's Avatar
 ShadowFox 
CO/USA
 
Experience: Intermediate
Platform: TradeStation, Multicharts
Trading: Stocks, Futures
Posts: 129 since Jun 2020
Thanks Given: 70
Thanks Received: 157

5/4/2021

Closed Trades: 4 $117 (including commissions)

Open Trades: $149.82
Long QM $375
Long YK $8.75
Short MES -$6.25
Short MGC -$16
Long YC -$51.25
Short MNQ -$160.5
Great trade in gold from yesterday closed out today. Only to get short today and give a little back (unrealized).


Not a great close to the crude trade. If the move would have started a few minutes earlier I would have stayed in the trade but the system said close. I did get back into QM later in the day for a nice run though. Not bad.

Visit my NexusFi Trade Journal Started this thread Reply With Quote
Thanked by:
  #8 (permalink)
 
ShadowFox's Avatar
 ShadowFox 
CO/USA
 
Experience: Intermediate
Platform: TradeStation, Multicharts
Trading: Stocks, Futures
Posts: 129 since Jun 2020
Thanks Given: 70
Thanks Received: 157

I can already see I am over invested in energy for the system. Sadly the E-Miny contracts aren't so mini. QM is only 1/2 a CL contract. Not really sure what to do about it at this point until I can scale everything else up. This could really throw off my drawdown requirements for proving the system. I am contemplating removing QM from the portfolio until the risk can be shared between the assets a bit better, which means more capital and more contracts. If one contract can carry or break the whole system then it is not worth the risk. Going to sleep on it.

**Update** Took profits on QM and turned off ($512 profit). I need to assess overall system performance using micro E-miny contracts instead of full size contracts. I like the overall diversification on full size contracts but QM is weighted 5:1 currently vs. the original systems intent. The risk is too great for the portfolio capital.

I know I broke a rule of systems trading by overriding the system but in this case it is okay. It is not good practice to override the system if you are going to continue to trade the system, but if the system is flawed and you need to reassess or completely change it, cut the trade and turn the strategy off.

The portfolio and system remains live and unchanged outside of QM. I will reassess overall performance expectations tomorrow with proper weighting on full size contracts to reflect scaled returns of each individual asset.

Visit my NexusFi Trade Journal Started this thread Reply With Quote
  #9 (permalink)
 
ShadowFox's Avatar
 ShadowFox 
CO/USA
 
Experience: Intermediate
Platform: TradeStation, Multicharts
Trading: Stocks, Futures
Posts: 129 since Jun 2020
Thanks Given: 70
Thanks Received: 157

So following up to decisions made yesterday. As stated in post 1, the portfolio and each individual asset is based on full size contracts. The plan was to trade micro/E-miny contracts to reduce the capital risk during live testing. The strategy runs on the full size contracts though, it just trades the micro/E-miny. The errors I made initially was expecting each micro/E-miny was somewhat equivalent scale vs full size. This was very wrong.

Financials - 1 NQ = 10 MNQ or 1:10 (same for ES and GC)
Energy - 1 CL = 2 QM or 1:2 (same for NG)
Grains - 1 C = 5 YC or 1:5 (same for W and S)

So when I made the assumption that I could divide the overall system drawdown by 10, I was very wrong. To get the equivalent of trading 1 micro/e-miny contract for the portfolio I need to scale up the Energy and Grain contracts to match. So if I were trading 1 MNQ, 1 QM and 1 YC, in order to replicate on full size, I would trade 1 NQ, 5 CL, and 2 C. big difference vs the original assumptions. Then I would divide the results by 10 to get the equivalent micro/E-miny conversion for drawdown/run-up, etc.

So here is a rundown of what that looks like

Original assumptions




Actual relationship to micro/e-miny, you can see the difference in drawdown here. Much larger risk than originally assumed.




CL or QM removed




CL and NG or QM and QN removed




You can see that my original assumptions of a max drawdown of ~$1800 was quite off when the scaling is accounted for. The new max drawdown is close to $4000 equivalent. This is a BIG difference. The plan is to trade the portfolio with CL and NQ removed due to the imbalances. I cannot add these without increasing the risk of the overall system and therefore increasing the capital.

Here is the short-term look at drawdown with CL and NG removed to show we are still in a decent normal drawdown to turn the system on.


Long explanation to say I royally messed up on my assumptions which justified the portfolio change. Live and learn

Visit my NexusFi Trade Journal Started this thread Reply With Quote
  #10 (permalink)
 
ShadowFox's Avatar
 ShadowFox 
CO/USA
 
Experience: Intermediate
Platform: TradeStation, Multicharts
Trading: Stocks, Futures
Posts: 129 since Jun 2020
Thanks Given: 70
Thanks Received: 157


One week down and things are shaping up. My thoughts on starting the strategy during a normal drawdown period were almost perfect. Had a nice run-up right after turning the system on then the drawdown double bottomed but has since run-up to almost a new all time high.

Still trading 1 MNQ, 1 MES, 1 MGC, 1 YW, 1 YC, 1 YK contract.

5/3/2021: Starting equity = $9,300
5/10/2021: Current real-time account equity = $10,101.29 (1 trade open)

Here is the updated short-term equity chart shared in the last post after updating for my new portfolio adjustments (divide totals by 10 to compare to my system as its based on full size contracts). Blue markings are when I went live with the strategy.

No issues with fills on micro contracts, everything is matching up nicely.

I don't want to get ahead of myself but the next step is a process for scaling up. Will be a few months from now but I want to get an idea of what that might look like while I continue to work on portfolio diversification.

I also need to work on sleeping and not worrying about the system. This is my first overnight system with markets that are open so that is a whole new element. I trade stocks overnight but I don't trade pre or post markets so nothing needs to be monitored overnight on those systems.


Visit my NexusFi Trade Journal Started this thread Reply With Quote
Thanked by:




Last Updated on March 1, 2022


© 2024 NexusFi™, s.a., All Rights Reserved.
Av Ricardo J. Alfaro, Century Tower, Panama City, Panama, Ph: +507 833-9432 (Panama and Intl), +1 888-312-3001 (USA and Canada)
All information is for educational use only and is not investment advice. There is a substantial risk of loss in trading commodity futures, stocks, options and foreign exchange products. Past performance is not indicative of future results.
About Us - Contact Us - Site Rules, Acceptable Use, and Terms and Conditions - Privacy Policy - Downloads - Top
no new posts