[KenHall]

I had my systems off for the last couple of months of the year while I helped a Tesla compete in the Australian leg of a recreation of the 1968 London to Sydney marathon, ran a marathon in bits across outback Australia raising funds for RFDS and then raced my first half Ironman triathlon in Dec after a big year of training.

System performance for the year appears to be as expected, but recognise I need to put some manual effort in taking account of spinoffs / mergers etc to accurately track monthly performance rather than just read from the Alera PNL stats.

CY2022 Stats
WTT ASX       -33.36%     
ASX Momo     2% 
US Day Trade L      4%
US Day Trade S     –4.1%
TLT US        -31.8%

Overall account performance for my main account for the year doesn’t look too bad (-12.5%) but this is due to not having all cash allocated or not making use of leverage appropriately to reduce counterparty risk.Speaking of which, I was caught up in the FTX debacle/fraud at the end of the year while galavanting around the country – I had been testing a long/short MR system on FTX and my withdrawals were not honoured before they went into administration! Expect it will be a long time if I ever see anything back from it. I got me thinking what I could do better to mange this type of risk.

Remove funds from brokers/exchange if not tradingRemove funds from brokers/exchange if regime filter is active
Minimise funds/counterparty risk by using provided leverage
Spread diversify funds/risk between multiple brokers/exchanges
Review if I even need to be exposed to the risk – see this excerpt titled “ENOUGH”  from Reducing the Risk of Black Swans by Larry Swedroe that I’ve reposted in the psychology section of the forum.

I’m not convinced the FTX contagion is finished yet (DCG / Genesis / Binance etc) so will stick to developing/backtesting crypto systems for the moment – no trading!

Moving on to 2023 the concept of a portfolio of loosely/un-correlated systems as presented in Laurens Bensdorp’s – Automated Stock Trading Systems appears attractive to me. My own limited research has confirmed the value of combining systems to smooth returns and minimise drawdowns. This will involve porting and testing systems with RealTest as it’s easy to see the impact of a portfolio approach compared to AB. I know a few on this forum have done this already and Nick himself has recently committed to doing this as well, so I’ll be in good company and there appears to be lots of help both here and in the realest forum.

Happy New Year to you all!

[KenHall]

June 2022 Stats
WTT ASX -0.7%
ASX Momo -1.9%
US Day Trade L -13.64%
US Day Trade S -0.1%
TLT US PP

 July 2022 Stats
WTT ASX -1.0%
ASX Momo 2.3%
US Day Trade L 0.5%
US Day Trade S -1.0%
TLT US PP

Had most of June off with a well overdue visit to family in Scotland which was awesome, so posting a double month update…

Been further developing/testing/evaluating the (intraday crypto futures) MOC style system I’ve been working on over the last few months…

Added a short side to the system based on regime flip – this appears to help to keep the equity curve heading in the right direction and reduces MaxDD but I’m not yet fully convinced as it’s not showing robustness against another similar market I’ve tested.

I’ve analysed actual trades vs Amibroker expected backtest results for 100 trades (2 months) – although there are some + and – differences in %profit and exit price (mkt orders) – on average it turns out they even out. Specifying that price has to go below my (long) limit order to consider a fill in AB, and using accurate symbol level TickSize and LotSizes has helped a lot with backtest vs live order placement and trading accuracy.

I’ve been trying out RealTest for the first time to help with analysing the profile of the MOC system. It’s trivial (in comparison to AB) to see the effect of multiple long/short stretch levels simply by defining each stretch as it’s own strategy within a RT script/portfolio. RT’s ‘Summary Report’  has been particularly useful to evaluate how each stretch level/strategy contributes to portfolio level ROR, MaxDD, no of trades and expectancy etc. An an aside, the quality of discussion and example scripts/strategies shared in the RealTest forum appears to be excellent.

I’d still like to test volatility based position sizing and a market breadth type regime measure before committing fully to the system.

And I’ve started to look at the merit of bringing the idea back to daily stocks – which will have some execution hurdles to overcome with IB – possibly needing to use Alera’s APM Conditional Order/OTL capability which allows you to ‘Create a new Limit Order’ once our initial order has been filled.

[KenHall]

May2022 Stats

WTT ASX -23.7% **
ASX Momo 2.3%
US Day Trade L 0.65%
US Day Trade S 0.32%
TLT US -0.2% Mostly PP

** WTT took some hits this month – in addition to being in the trading halt of AVZ Minerals we had the TAH (-80%) spinoff of TLC. It’s not as bad as it looks with TAH as TLC wasn’t yet recognised in the portfolio, but will be next month. 

Feeling comfortable about being in mostly cash now and it’ll be interesting for sure to see how things pan out and what I learn if we have a sustained bear market / recession. It’ll be my first while trading Equities!

[KenHall]

April 2022 Stats

WTT ASX -1.0%
ASX Momo 0.0% PP
US Day Trade L -7.65%
US Day Trade S 0.0%
TLT US -4.6%

Had some operational issues with APM / IB / Speedy on Fri 22nd which meant I kept a load of MOC positions over the weekend, this time it cost me! Otherwise Day Trade Long would have been neutral for the month.

Need to be a more attentive to reviewing things daily!

[KenHall]

RE: The only other problem that I have been noticing lately is that I have had a few instances of trades being in my exploration and entered in real trading yet the next day they do not appear in the backtest and a different stock might. Has anyone else come across this?

Hi Ben, I’ve had some strange issues like this when developing an overly complicated system that relied on identifying Higher Highs HH, Higher Lows HL etc that I’ve since shelved. (I’ve had trades not taken in real life (but in the backtester), but that’s normally because I had a buy limit order at the Low of the candle.)

I’d encourage you to get to the bottom of the issue as it may indicate an error that invalidates the confidence you have in the strategy.

Couple of suggestions…

Check for any possible post-dictive (forward looking, signal repainting) errors especially in your ranking code

Check for any data integrity inconsistencies.

Use the “Interpretation Window” (Course: module 10), to visually step through the bars (and with bar-replay) and see that indicator results, ranking results are what you expect.

Use the bar-replay feature to step through each day generating explorations / backtests (with _TRACE debugging) to see if it’s consistent in creating the same signals / backtest results. (Initially I never knew that bar-replay essentially truncates your database to the time period at the bar-replay, therefore Amibroker can not see future data and so best simulates what happened on the actual day)

Reach out to Craig to review your code.

[KenHall]

March 2022 Stats

WTT ASX 5.7%
ASX Momo 0% PP
US Day Trade L 2.7%
US Day Trade S -1.4%
TLT US -0.4% Mostly PP

[KenHall]

I had 4 MOC orders get canceled for some reason last night and so now still in those positions…

Looks like IB had some connectivity issues last night that many are reporting – https://twitter.com/brent_calver/status/1501254073365917700?s=20&t=-3CVvQLfxQgOIO1hnRGX_Qhttps://twitter.com/brent_calver/status/1501254073365917700 TWS/IBGW and the IB app was unresponsive for at least 30 mins…

My APM logs show I had lost connectivity to IB via TWS/IBGW about 10:45 and 11:45 EST.

Still not sure why my 4 MOC orders got cancelled though – other got through fine in the session.

Code:

[8/03/2022 10:26:20 AM][1004] [ INFO] [ConditionalOrder.SLB] Activating RSN order
[8/03/2022 10:26:20 AM][1004] [ INFO] [ Signal.SLB] BTC SLB MOC
[8/03/2022 10:26:20 AM][1004] [ INFO] [ PlaceOrder.SLB] BUY SLB ### SHARES MOC – – – DAY
[8/03/2022 10:26:20 AM][1004] [ INFO] [ Order.SLB] Sending order 21475 to IB

[8/03/2022 9:50:18 AM][1004] [ INFO] [ConditionalOrder.XOM] Activating RSN order
[8/03/2022 9:50:18 AM][1004] [ INFO] [ Signal.XOM] BTC XOM MOC
[8/03/2022 9:50:18 AM][1004] [ INFO] [ PlaceOrder.XOM] BUY XOM ### SHARES MOC – – – DAY
[8/03/2022 9:50:18 AM][1004] [ INFO] [ Order.XOM] Sending order 21473 to IB

[8/03/2022 10:26:20 AM][1004] [ INFO] [ConditionalOrder.SLB] Activating RSN order
[8/03/2022 10:26:20 AM][1004] [ INFO] [ Signal.SLB] BTC SLB MOC
[8/03/2022 10:26:20 AM][1004] [ INFO] [ PlaceOrder.SLB] BUY SLB ### SHARES MOC – – – DAY
[8/03/2022 10:26:20 AM][1004] [ INFO] [ Order.SLB] Sending order 21475 to IB

[8/03/2022 10:45:02 AM][—-] [ ERROR] [ EW CodeErr] -1:2157:Sec-def data farm connection is broken:secdefhk
[8/03/2022 10:45:03 AM][—-] [ ERROR] [ EW CodeErr] -1:2158:Sec-def data farm connection is OK:secdefhk
[8/03/2022 10:46:17 AM][—-] [ ERROR] [ EW CodeErr] -1:2157:Sec-def data farm connection is broken:secdefhk
[8/03/2022 10:46:18 AM][—-] [ ERROR] [ EW CodeErr] -1:2158:Sec-def data farm connection is OK:secdefhk

[8/03/2022 11:30:42 AM][1005] [ INFO] [ConditionalOrder.UNH] Activating RSN order
[8/03/2022 11:30:42 AM][1005] [ INFO] [ Signal.UNH] STC UNH MOC
[8/03/2022 11:30:42 AM][1005] [ INFO] [ PlaceOrder.UNH] SELL UNH ### SHARES MOC – – – DAY
[8/03/2022 11:30:42 AM][1005] [ INFO] [ Order.UNH] Sending order 21479 to IB

[8/03/2022 11:32:03 AM][1005] [ INFO] [ Signal.RHI] STC RHI MOC
[8/03/2022 11:32:03 AM][1005] [ INFO] [ PlaceOrder.RHI] SELL RHI ### SHARES MOC – – – DAY
[8/03/2022 11:32:03 AM][1005] [ INFO] [ Order.RHI] Sending order 21481 to IB
[8/03/2022 11:38:48 AM][—-] [ INFO] [ EW Canceled] Order 21481 cancelled
[8/03/2022 11:38:48 AM][—-] [ INFO] [ EW Canceled] Order 21475 cancelled
[8/03/2022 11:45:44 AM][—-] [ ERROR] [ EW CodeErr] -1:2103:Market data farm connection is broken:hfarm
[8/03/2022 11:46:43 AM][1005] [ INFO] [ConditionalOrder.HSIC] Activating RSN order
[8/03/2022 11:46:43 AM][1005] [ INFO] [ Signal.HSIC] STC HSIC MOC
[8/03/2022 11:46:43 AM][1005] [ INFO] [ PlaceOrder.HSIC] SELL HSIC ### SHARES MOC – – – DAY
[8/03/2022 11:46:43 AM][1005] [ INFO] [ Order.HSIC] Sending order 21482 to IB
[8/03/2022 12:23:35 PM][—-] [ INFO] [ EW Canceled] Order 21479 cancelled
[8/03/2022 12:23:35 PM][—-] [ INFO] [ EW Canceled] Order 21473 cancelled
[8/03/2022 3:00:00 PM][—-] [ INFO] [ SendEmail] Sending account email
[8/03/2022 3:30:00 PM][—-] [ INFO] [ EW Canceled] Order 21441 cancelled
[8/03/2022 3:30:00 PM][—-] [ INFO] [ EW Canceled] Order 21449 cancelled

[KenHall]

Feb 2022 Stats

WTT ASX -1.3%
ASX Momo 7.7%
US Day Trade L 10.5%
US Day Trade S -1.5%
TLT US -0.1%

Mostly in capital protect mode now across the strategies with a few positions still live in the WTT.

The MOC long system still doing it’s thing well with another goof day bringing in most of the return for Feb.

[KenHall]

Jan 2022 Stats

WTT ASX -17.6%
ASX Momo -6.1%
US Day Trade L 21.6%
US Day Trade S -0.6%
TLT US -20.0%

[KenHall]

2021 Stats (From March 2021)

WTT ASX 32.5%
ASX MOMO -19.3%
US Day Trade L 4.0%
US Day Trade S -0.3%
TLT US 20.77%

Year in Review

It’s been great to see how the (WTT) trend following system jumps on and off a trend and also start to get an appreciation for mean reversion with the Day Trade MOC systems across both stocks and crypto. I now feel very comfortable with the style and running of the different types of systems.

I’m doubling the position size (now 10%) for the MOC systems for 2022, now that I’ve ran it for a year and have a bit of a feel for what to expect. I’ve also changed to the tiered pricing model with IB.

I’ve learnt loads through the year in automating my stock/crypto systems, using both Explorations and the the CBT in Amibroker alongside Alera Portfolio Manager (APM).

I’ve still got some work do on my crypto MOC/MR system, as there isn’t really such a thing as Market On Close in crypto – I need to do some work on analysing the effect of slippage and partial fills on these markets – I’m learning that with short term systems this could have a material impact on returns. Also exploring suitable regime or market breadth filter for this system.

[KenHall]

Kens Journal – Nov 2021

Monthly Returns

WTT ASX -3.2%
ASX Momo 2.1%
US Day Trade L -2.1%
US Day Trade S -0.3%
TLT US -8.36%

ITDL FTX -6.84%

Intra Day-Trade MR for Crypto (ITDL FTX)

Was green until the last couple of days of the month when I got spanked again!

I’m exploring limiting the no of fills I’ll take with this system to reduce risk (no of fills) when there is a market wide pullback.

So I’ll enter 20 limit orders for the period, but may limit the number of fills I’ll take to say 10.

Operationally, I can do this using Alera APM’s SoftCap feature – it’ll allow you to place as many limit orders as you like but will cancel (the others) as soon as you reach the cap.

But of course, I need to be able to backtest it – without selection bias.

Although I rank my limit orders, the market decides which limit orders get hit first. And I suspect that the first to get hit in a market wide pullback could be the weakest – i.e. those less likely to spring-back first!

The only way I can currently see of backtesting this properly is to use a more granular timeframe (5mins instead of 4H), similar to how we use daily bars for weekly, monthly rotational systems. And actually see which limit orders get hit first.

Probably over thinking this – as backtesting YTD maxDD is 16% across 3,600 trades!

Should (and likely to) split capital and combine with another MR strategy instead!

[KenHall]

Kens Journal – Oct 2021

TurnKey’s Monthly Returns

WTT ASX 6.35%
ASX Momo -6.48%
US Day Trade L 2.07%
US Day Trade S -0.27%
TLT US -5.08%

Noticed an issue with how I was running the weekly ‘Exploration’ for new buy/sell signals on WTT recently – basically I had only been looking at the exploration results and I think I have a common setting (Turnover for example) with the Day Trade systems that also run on the same speedy server/amibroker instance.

Long story short I should have closed some positions earlier than I did, lucked out as I was still in a position with INR that done well.

Lesson learnt – either hard code parameter settings in production systems, or use unique parameter names/variables!

Day-Trade MR for Crypto

Learning lots by running the day-trade system intraday on crypto… Found a way to query the exchanges for symbol TickSize data and import ito Amibroker to assist with Limit Order placement and backtest accuracy.

Practically learning about the dark side of Mean Reversion by getting spanked a couple of times recently – so been looking into what sort of historical daily movements I should expect based on positions sizing etc.

Correlation is really high with these markets, so looking at exploring a couple of the ideas shared in this post to smooth things out a bit.

https://blog.sutherlandresearch.com/index.php/2017/06/20/taming-mean-reversions-left-tail-dont-use-stop-losses/

[KenHall]

Ohh, and I found the course syllabus for you to review – https://enlightenedstocktrading.com/online-cryptocurrency-course-the-crypto-success-system/

[KenHall]

Hi Rob, I use https://sites.google.com/site/amibrokerplugins/tradingview_com_realtime_amibroker_data_plugin for data. Interestingly you can get data straight from the source (the exchange) using AmiQuote’s new javascript capability and have recently switched over to this source.

See

Binance: https://amibroker.com/members/library/detail.php?id=1558
FTX: https://amibroker.com/members/library/detail.php?id=1557

Binance provide an excellent historical source of data as well @ https://github.com/binance/binance-public-data/

[KenHall]

Hi Howard, apoligies for not noticing this sooner – reach out to Adrian Reid at https://enlightenedstocktrading.com ([email protected]) – easiest way is to send “Adrian Reid” a message via Facebook and mention your interested in understanding a bit more about “The Crypto Success System”

[Author]

Posts