stockmarket prediction ($QQQ) for Tuesday (October 23, 2018)
stockmarket prediction ($QQQ) for Tuesday (October 23, 2018)

I built an #AI to predict the $QQQ (tracks Nasdaq-100). It should work about 2/3 of the time.

"...wide-ranging, bullish-to-bearish, with a probable bullish high and a probable bearish low. Caution is somewhat warranted."

Monday met 0 of the 2 range(s) strongly predicted 2 day(s) before:

New people: welcome! You may need to acquaint yourself with basic terms like median and interval predictions. Ask if something is unclear - and either another follower or I can assist you. But please exercise your own due diligence as a first step.

Note: All numeric data (medians, intervals, and probabilities) are checked at least twice before posting. Other content is subject to a first review before posting in Patreon. The same content is then subject to a second review, and edited accordingly in Patreon, just before posting in Twitter. Major edits or errors are notified to patrons if they occur; minor ones (typos, etc.) are not.

stockmarket prediction ($QQQ) for Tuesday (October 23, 2018)


Tuesday promises a market that is wide-ranging, bullish-to-bearish, with a probable bullish high and a probable bearish low. Caution is somewhat warranted.

Absent exogenous events, the low may exhibit probable AND moderate bearishness.

The forecast has mostly a good performance by the primary, and some support by the secondary.

Tuesday can be traded in typical fashion by both bulls and bears - with the former going long in the low interval, and the latter going short in the high interval. Both should consider that the open is weakly-predicted, so there is more uncertainty surrounding the beginning of the session. A wait-and-see approach seems to be the better way.



1. All predictions are considered "strong" unless stated otherwise.

2. When applicable, the first number is the median prediction, followed by the range prediction based on a 95% confidence interval.

3. ! indicates "it is not the case that..."

open: >174.48; WEAK; primary is flat; secondary is bullish

high: 176.21; 174.93-177.49; primary is bullish; secondary is non-bearish

low: 170.97; 169.85-172.10; primary is bearish; secondary is non-bullish

close: !(173.34-174.48); primary is non-flat; secondary is bearish

(Keep in mind the last close was 173.91)


Note this forecast is quite similar to that for the previous trading day (October 22, 2018). Evidently, the medians, intervals, and probabilities are different.

The open is predicted flat by the primary at .52 probability, which is much below my .60 informal threshold for strong predictions. The secondary is bullish at .55, somewhat better but still too low. So, I consider the open to be weakly-predicted and thus too uncertain. That said, I will *slightly* suggest a bullish open.

(Note: A "non-flat" prediction is for the price to be outside the flat-band, which is the previous day's closing price +/- 0.33 percent. A flat prediction is for the price to be *within* the same band.)

The high is predicted bullish by the primary, and its probability is .61. This probability is good.

The secondary is non-bearish at .53, a negligible probability, even if permissive of the primary's prediction. In all, I opt for the primary's whole interval, without qualification.

The low is predicted bearish by the primary at .61. This probability is good.

The secondary is non-bullish at .60, which is permissive of the primary's prediction. In all, I opt for the primary's whole interval, without qualification.

Finally, as for the close, the primary predicts non-flat at .58. The secondary is bearish at .53. Even though the close-by-primary's probability is below my .60 threshold for strong predictions, I will override it here and opt for its non-flat close. I do this based on 3 things:

1. There is a small margin of 2 percentage points from the .60 threshold, which gives me some leeway to exercise judgment.

2. The close-by-primary has a relatively fair performance in the past (and last week, actually).

3. Other output (not shared) confirms a non-flat close.

So, it's sensible to adopt the close-by-primary's prediction of a non-flat close as a strong one.

Overall, the forecast quality is moderate. The primary has good high and low predictions, which are the backbone of the daily forecasts. Their intervals are fairly narrow and their probabilities are decent. However, the open is obviously weakly-predicted, which hinders the overall forecast quality.

The secondary provides some support. It is somewhat informative on the open and low, but not quite useful elsewhere.

The tradeability of this forecast is moderate-to-high.

In general, Tuesday should provide the typical opportunities.

For bulls, you can go long from the low interval and aim for the high interval. For bears, you can go short from the high interval and aim for the low interval.

Do exercise your own judgment in being conservative with any gains: tighten stops or close positions as you see fit, so you don't have to wait for both intervals to be met successfully.

For example, if you are bearish and shorted on the high interval, it's ok to get out in the flat-band or just below it. A profit is a profit. Of course, the higher the probabilities for a given prediction, the more risk you are willing to assume. But Tuesday we have strong, yet not too strong, probabilities (in the 60s). This should be a point of caution in your trading.

At the same time, there is reassurance in the strong prediction of a non-flat close. If you are trading in the late afternoon, and the QQQ is already inside an interval, you can sensibly expect it *not* to return to the flat-band by day's end. This can give you hope if you want to push for further gains.

Also, remember that Elixir does NOT sequence the high and low predictions - so, we don't know which one will come first. The day could easily be bearish for most of the session, fulfilling the low prediction, and recover to meet the high prediction by the end. This is why I recommend, say, that you go short inside the high interval. It increases your chances of having a good "probabilistic ceiling" from which to ride the market down.

Interval players are agnostic, so they can play both bullish and bearish strategies.

As a whole, I would let the market take the first step, and then use the intervals to map out the upcoming probabilities. Take into account that the intervals are like bell-curves of probabilities, in which the median surrounds the most probable outcome - and either tail of the curve is more improbable.

Finally, if we have negative exogenous (bad news) in the morning, expect the market to likely go towards the low interval - while ignoring the high one. If the exogenous is positive, then the opposite is likely - a run towards the high interval, but a disregard of the low interval. You can then try to play whichever interval is engaged.

Good luck. Keep it karmic and spread the love. Besides helping myself, I want to create a public good for everyone outside Wall Street.

If you are interested, this link takes you to previous week-in-review's, so you can verify our past performance:

If you want to get early predictions for trading/investing, consider becoming a patron. It's $10/month, and helps support my AI-lab:

If you are international, that's fine - Patreon accepts Paypal.

Also, I chose $10 because it's an affordable amount for people beyond the US. (And also inside the US!) Besides helping myself, I want to democratize access to good financial forecasting.

See you later.

predictionmonk released this post 1 day early for patrons.   Become a patron
By becoming a patron, you'll instantly unlock access to 132 exclusive posts
By becoming a patron, you'll instantly unlock access to 132 exclusive posts