andyqi2
01-23-2016,
Hi, guys/gals. I am a retired economist snowbirding between Florida and Ottawa, with lots of data and some self taught programming experience, mostly spreadsheet macros before getting into Easylanguage and Ninja. Was working with Omega Supercharts in my 20s but gave up, focused more on real estate, returned to Tradestation couple years back now that I can work at this full time, and being inspired by a Florida neighbour who actually makes a living from trading, but migrated to Ninjatrader (https://futures.io/wiki/trading-wiki/Ninjatrader) for the bulk backtesting (https://futures.io/wiki/trading-wiki/Backtesting) in the basic package and price (free!).
To my knowledge, there is no way to do direct algo (equities) trading from within a Canadian RRSP (our version of an IRA (https://futures.io/wiki/trading-wiki/IRA)), so the licensing fee is moot. NinjaT needs to partner with at least one Canadian equity brokerage that does RRSPs, and I think Virtual Broker is the obvious choice. Or better yet, open a Canadian branch of NinjaT Brokerage! I don't want to bust open my RSP money and open a futures account until this becomes more serious and I get the tax implications straight in my mind.
My current thinking is that if I ever start making money, trading in a regular account in Canada will be taxed as capital gains, whereas any profits I build up in my RSP will eventually come out at top marginal rate, so trading outside an RSP may make more sense. My understanding is that Canadian tax rules on day trading are very different than US, so IRS rules are not a guide. Comments?
Futures such as ES are not permitted in any RRSP, and IB (the only Ninja-capable equities broker in Canada) does not have RSP accounts. I also need a bit more trading experience to qualify for an IB account. I must confess, I don't like the IRS disclosure paperwork involved in opening a Canadian IB account, but its not a deal breaker. Love that IB founder, especially his 2012 political ads attacking socialism.
My current workaround is to trade Dow and SP500 3x ETFs (https://futures.io/wiki/trading-wiki/ETFs), UPRO and SPXU seem to currently work best in backtesting, sometimes using the signals/movement from the ES or the ES as confirmation. I am getting pretty good at finding patterns that repeat and walk forward well for a few months, but get concerned when patterns will not walk forward in previous seasons or years. Becoming aware that one needs to treat each month of the year as a different market?
My current approach, the last couple of weeks, is to abandon trying to find a pattern that walks forward with a good profit factor going a full year back in history, and rather use my backtested system as an indicator, but pilot with hands on discretion, trading 10 shares. A good day I am up $2, and bad day I am down $2, real money after the $1 commission Virtual Brokers charges. On my small sample, I am up most of the time. I am surprised at the quality of my fills using odd lots.
I find I need the speed/action of 1 minute bars to hold my attention. I use CCI to early-signal overbought/oversold, inspired by a futures.io (formerly BMT) video on CCI support/resistance (https://futures.io/wiki/trading-wiki/Support-Resistance) charting, and a slightly slower composite indicator I developed to signal the beginning/confirmation of retracement, both optimized on length and trigger level (degree of deviation from mean to start paying attention to), paying attention to which test versus optimize period lengths walk forward best.
Though with discretion, I am now more just looking at the raw level of the indicator and using backtesting to inform how much respect versus caution to give for a given level. I also optimize on period of day, which suggests open to 11 am or so is worthwhile, and maybe the last hour. I look at day of week, but don't have the discipline to not trade marginal days indicated in back test (https://futures.io/wiki/trading-wiki/back-test), without confirming that these are untradable with discretion.
I look to get out of a trade when there are a couple of violations of the move. I ignore bar tails, I even make them invisible n the chart, and if a bar opens or closes below the recent low, that's a first violation, and then if a second bar opens or closes down, then I am done. I am ever mindful of the occasional flash crash that comes through like a hurricane every few years. My neighbour mentor who came up out of the Chicago pits got burned more than once holding long positions and will almost never trade long...he only trades short..."markets never crash up", he says.
To my knowledge, there is no way to do direct algo (equities) trading from within a Canadian RRSP (our version of an IRA (https://futures.io/wiki/trading-wiki/IRA)), so the licensing fee is moot. NinjaT needs to partner with at least one Canadian equity brokerage that does RRSPs, and I think Virtual Broker is the obvious choice. Or better yet, open a Canadian branch of NinjaT Brokerage! I don't want to bust open my RSP money and open a futures account until this becomes more serious and I get the tax implications straight in my mind.
My current thinking is that if I ever start making money, trading in a regular account in Canada will be taxed as capital gains, whereas any profits I build up in my RSP will eventually come out at top marginal rate, so trading outside an RSP may make more sense. My understanding is that Canadian tax rules on day trading are very different than US, so IRS rules are not a guide. Comments?
Futures such as ES are not permitted in any RRSP, and IB (the only Ninja-capable equities broker in Canada) does not have RSP accounts. I also need a bit more trading experience to qualify for an IB account. I must confess, I don't like the IRS disclosure paperwork involved in opening a Canadian IB account, but its not a deal breaker. Love that IB founder, especially his 2012 political ads attacking socialism.
My current workaround is to trade Dow and SP500 3x ETFs (https://futures.io/wiki/trading-wiki/ETFs), UPRO and SPXU seem to currently work best in backtesting, sometimes using the signals/movement from the ES or the ES as confirmation. I am getting pretty good at finding patterns that repeat and walk forward well for a few months, but get concerned when patterns will not walk forward in previous seasons or years. Becoming aware that one needs to treat each month of the year as a different market?
My current approach, the last couple of weeks, is to abandon trying to find a pattern that walks forward with a good profit factor going a full year back in history, and rather use my backtested system as an indicator, but pilot with hands on discretion, trading 10 shares. A good day I am up $2, and bad day I am down $2, real money after the $1 commission Virtual Brokers charges. On my small sample, I am up most of the time. I am surprised at the quality of my fills using odd lots.
I find I need the speed/action of 1 minute bars to hold my attention. I use CCI to early-signal overbought/oversold, inspired by a futures.io (formerly BMT) video on CCI support/resistance (https://futures.io/wiki/trading-wiki/Support-Resistance) charting, and a slightly slower composite indicator I developed to signal the beginning/confirmation of retracement, both optimized on length and trigger level (degree of deviation from mean to start paying attention to), paying attention to which test versus optimize period lengths walk forward best.
Though with discretion, I am now more just looking at the raw level of the indicator and using backtesting to inform how much respect versus caution to give for a given level. I also optimize on period of day, which suggests open to 11 am or so is worthwhile, and maybe the last hour. I look at day of week, but don't have the discipline to not trade marginal days indicated in back test (https://futures.io/wiki/trading-wiki/back-test), without confirming that these are untradable with discretion.
I look to get out of a trade when there are a couple of violations of the move. I ignore bar tails, I even make them invisible n the chart, and if a bar opens or closes below the recent low, that's a first violation, and then if a second bar opens or closes down, then I am done. I am ever mindful of the occasional flash crash that comes through like a hurricane every few years. My neighbour mentor who came up out of the Chicago pits got burned more than once holding long positions and will almost never trade long...he only trades short..."markets never crash up", he says.