Currency Scan

A quick look at the current price action. Some of this is residual from Friday. Technical and ML models are mostly out of sync, I'll be checking things again before the US sessions kicks off.

A couple of things showing up as trades, neither of which I'm taking right now. The USDCAD is so blown out I want to wait and let my chart technicals catch up. It's also screwing up my position sizing as the current forecast is for a move up but the price is currently below the level I'd use as a stop... All signs point to "wait".

Yesterday morning I jumped into the AUDCHF, CADCHF, EURAUD, EURGBP and the GBPUSD. Things were pretty quiet all day, not a lot of movement in the account.



This morning on the rebalance I added NZDJPY and USDCAD out of a handful of new signals (trying to keep the correlation somewhat in check).



The CADCHF signal actually fell off the grid this morning as things have gotten mixed in both the technical measurements and the ML forecasts. I could pull it, not sure what to do. I basically attack these exits from SL, TP and Time. On top of that, there is still a compelling chart setup to consider. The thing is riding on top of its trend channel and we often see a reversal from there. Probably just a coin flip at this point. I'm inclined to let it play out

Shorted USDCHF @ 0.9969
Last night I shorted the USDCHF around 0.9969, the trade was relatively flat through the early trading and is starting to move against me a little bit here in the early US session. I went in about half the recommended lots as I was expecting some additional stuff to hit the scan and wanted to leave room in the risk profile. Sort of glad I did.



Heartburn ahead
Part of the signal on the pair has dropped off which throws me into a little bit of no-mans land. I still have supporting forecasts in the machine learning models so for right now I'm going to stick with it. My expectation is the pair will push against me a bit and we'll see the signal strength improve, at which point if all the stars align I'll add to the position a little. At that point, I'll have done about all I can and will just have to let it play out.



The most recent scan is signaling short in both the USDCAD and the NZDJPY. Signal strength is really low (position sizing is adjusted accordingly as a result) but I'm not ready to pull the trigger on it or the NZD just yet. Next scan runs in about fifteen minutes, will reassess then.

On a different note, I think (hope), I'm finally starting to get the trading dialed back in. December I drew down around two percent getting the risk model sorted and the first part of January was listless as I focused on intra-day trades. I just don't think ML is any good there, too much noise and I felt like all I was doing was generating spread revenue for the broker. As much as I hate holding trades over-night, especially in a small account (black swans and all that), not sure I'm going to have a choice. My model efficiency will probably pick up the further out I forecast. It's just a matter of surviving volatility associated with those long time frames. At any rate, I feel like we're getting the engine tuned up and account performance going forward will be tied into model accuracy, not poor risk modeling or inappropriate windows.

USDCAD Neutral NZDJPY getting bullish into a short call

The USDCAD has fallen off the scan and I'm not surprised, signal strength was really low and there was bearish momentum already in place. The NZDJPY is still signaling short but there is some bullish momentum in the current price action, so I'll hold off until the next scan to reassess, I'll probably be in this thing before the day is over.

NZDJPY testing the upper oscillator band. It has room to run to the upper trend line if momentum persists. I'd be just as happy to see it push up there before shorting it (may not have the luxury) but I'd feel better as the technical signal strength would be better. That's it for now, will be running things again in a couple of hours and will update when I do.

Closed USDCHF

I've closed the USDCHF, there is some bearish momentum in favor of leaving it open, but I'm also looking at a 4-hour forecast (not terribly reliable) with potential for stalling or a slight reversal, so I'm just gonna call it good for now.

Currently Short NZDJPY and USDCAD

Signal strength is low in both which typically indicates it's going to be a long grind. Going in relatively light so there is room to add in case the timing is completely off.

There are also a few other things in the watchlist to keep an eye on, EURGBP, GBPCHF, GBPJPY and GBPUSD (feels like a theme there) all have good signal strength in the asymmetry model. The rest is a little mixed, either the technical models are out of gear or the machine learning models are out of gear. My risk profile doesn't allow for jumping in when it's like that but those could warrant some additional research if people are hunting for an opportunity. That's a recommendation to research, not trade. Do your homework.

Closed the USDCAD and NZDJPY, Currently Flat

Those two trades ended up in pretty tight negative correlation, with no ground in the account being made as they gyrated around. That is, of course, better than having them lock step burning against me, so I'll take it. No harm done. My gut is, the NZDJPY could continue on its southernly path and closing it out may have been premature, but signal strength on my end was all but gone and the CAD had gone into "mixed" mode, so killing them was the only option I had at this point. Going forward it may make sense to build something of a meta-model to establish the probability associated with open trades. I've got to kick that one around a little.

Dollar Review

Some of your best trades are the ones you don't take. Last night I had three USD pairs signaling short (AUD, EUR, and NZD). The signal in each of them was really week and the potential for being on the wrong side higher. We were also sitting on the Fed in less than 24 hours so I just sat on the side. Glad I did.



About an hour before the Fed announcement (11:00 AM my time) the short term, 4 hr. model switched gears to "long" on all three trades. That particular model isn't very accurate but it was curious to see it turn in all three pairs. I wasn't in the market, but if I had been, I think I would have started sweating just a bit right there.



Apparently, the market had priced in something a little more aggressive than "patient" by the fed and the dollar took a dive as a result. The question now is, has it oversold? If NFP comes in stronger than expected Friday, we could see a sharp reversal. I'll probably sit this out as well. There is nothing in here at the moment compelling enough to risk the volatility. See you next week.

AUDUSD was hot for me and got really good returns

Nice! Back in the day I probably would have caught some of that on a breakout or momentum style entry. Now I'm just too lazy to watch the market, lol.

I actually entered the pair late thus did not catch the first news spike. However, I was contrarian and placed trade accordingly. The FOMC on the USD spike I also not in the mix.

If anyone is looking for a decent, independent data source, I've been using https://eodhistoricaldata.com/ for my back-end. The API is straight forward, the only downside I see is there isn't quite as much flexibility in the time intervals. Subscription is reasonable and it gives you access to FX and Global Equities, which is really nice if you're also trading stocks or keeping on eye in the broader indices. For most of us broker data is fine, I just like to have a back-up and thought I'd pass it along.