Mark, we had this argument 15 years ago, check your posts, I have excellent memory of you saying just the opposite.... that you had to know whether you were in a bull or bear market for longer term decision-making.
So you have changed your mind, have you?
Which Mark???
..well exactly.
If it was me - heck, 15 years ago..I change my mind on stuff ..I think its called 'learning/experience'...i've learnt ALOT in the last 15 years and changed my mind on many things that is for sure.
I have bit more time so let me say, I believe Mark(OEX) has a very good sense of the market, which is maybe something that is not totally quantifiable in TA .we could loosely call it 'read of the tape' which
is very experienced based. Whether that is partly due to knowing what the main trend is - bull/bear ... partly but I see in my own evolution that is partly the 'school of market hard knocks' type deal -learning what is
important and what isn't - and I think, some of this is very hard to communicate to others, so, we may say things differently and kind of argue semantics but end up actually thinking similarly.
So what I was trying to say in that post about prediction, was more about avoiding being dogmatic and rigid in belief about a certain trend ...more than saying its not useful to know the trend - obviously its very useful to
know if it a bull or bear! ...but my point is, we cannot know for sure - and maybe best we can do is 60-80% 'sureness' .. and so we must remain flexible to change that view based ideally on some method that shows
good backtested history.
Right now If I was forced to answer i've said my 'math' says its a bear market - and turns to bull if we get a weekly close above 4300 ....but my main goals is to try to be in synch with the one trend below that but it changes,
sometimes that trend is also 'murky' so as I said
->its about trying to synch with the clearest trend you can identify, whatever that is!
hope this is clearer.
I figured we were not as far apart as one might think.
I wanted to add that, for me, there's a huge psychological benefit to knowing if we are likely to be trading like a Bull or a Bear Market. My stuff works very differently in a Bear Market, and in point of fact some very good indicators (especially the Buy signals) don't work so well. Now, I know we're supposed to be sang froid as traders, but for me (and I think most of my subscribers), the psychological cost of a long string of losers--even if they are small--is remarkably high, often resulting in not trading signals that actually should be taken when they should be taken.
Avoiding those frustrating runs is really beneficial and not just avoided losses but also missed bull moves.
The Other Mark
Yep - both great points Mark ( ones I bolded).
My stuff is a little different on the first point -
- The Model I developed 'in theory' (and we all know how well 'theory' works :-) ) ..is independent of bull/bear for the IT signals. It a big topic of how - but like I say it uses 'math' of 'cause/effect' Reversion to mean/extreme, I would say its
actually worked to be fairly independently - but I still use TA to add 'sanity check' - I want the Model signals and the TA to AGREE to get high confidence.
For TA I do definitely agree with you, it works differently in CONTEXT of bull or bear, as over the years i've made this point many times that
> is often more important NOt what the signals say 'buy/sell' but HoW the market REACTS -
So - If I have a high odds winning BUY, and it doesn't work out ( either flat or down market) - then as you say, its an indication the next higher trend is DOWN ( so for IT signals, would mean Bear market).
It's funny - this is one of those experience things, that, its 'obvious' ..but somehow is hard to follow, and I would say suffers from 'selection bias'
- when a signal 'doesn't do what it should' - we can choose to another signal and say 'well, I should have used this one' - there are ALWAYS some signals that are 'right', and 'wrong'.
I suffered from ^this ( and many other trading/TA learning curve issues ) - I had so many signals, so could always kid myself 'you should have looked at this not that'.
( I said this was a long discussion I was trying to avoid LOL).
>>My SOLUTION in the end - was to PICK only a smaller number of 'the best of best' TA signals - and ignore the rest, because at that point, all they do is 'confuse the issue'.
If some asked me at this point, i'm at 23 years'ish and still learning ... I would probably say Just look at Internals - summations/NYMO in terms of KISS.
Ideally, I think you also need Sentiment - we both agree on that being key - and we could lump VIX into that, though, I think its more than sentiment and I use it a key part of my model.
Your second point is well taken - regarding effect of losing streaks on either missing the next good signals, or could be reducing position size too small.
This is where ALGO or mechanical systems have a big advantage!
Part of the problem is often feeling the need to 'always have a position' - this gets back to my point about prediction
- We do NOT have to be LONG or for that matter short - we can be in CASH as a 'position'
- I learnt this day trading really - that my job was to 'stalk the optimal setup' - where RIsk reward and win/loss probability peaks ...again humans (not just me I think) SUCK at that, having the patience to WAIT for that
moment like a predator catching prey .... as humans we also have pesky issue like losing mental focus, getting tired, having distractions etc which I why I stopped day trading and moved now to longer timeframes where
its not such an issue.
Edited by EntropyModel, 25 February 2023 - 06:21 PM.