If Technical Analysis doesn’t work, how do you define a bull market?

[longandshort]

I think TA works well enough. My query was why anyone who thought it didn't could talk about a bull market, which to me pre-supposes some implicit TA.

A bull market just means prices are rising. It’s not a prediction about the future.

Markets are semi-strong firm efficient, which means very little excess returns from analyzing historical data (technical OR fundamental!). Modern technical analysis used at CTAs and hedge funds revolves around the quantitative analysis of trends (momentum factor) and large data sets. The charting methods of the old days (40s to 70s) are largely eroded, and just the same with historical financial analysis has diminished. (Graham himself said that there was no more money in the strategy he articulated in the Intelligent Investor). Neither technical or fundamental analysis are good enough.

To generate excess returns you must exploit flaws in behavior, informational edge, or analytical depth.

 

[Fonz]

If you are an investor, perhaps just like that:

 

[smallfil]

TA doesn't work. In what way doesn't it work? How are you using it? It's like saying a hammer doesn't work.

Fundamental analysis doesn't work for me.
It gives me a bias that keeps me in losing trades and ties up my capital.
Accountants use GAAP (legal fraud) to publish numbers that the public has access to.
Everyone interprets the numbers differently.

There was an accounting scandal then, remember Enron? Top accounting firms were exposed just accepting the figures given by companies instead, of doing their jobs and auditing the financial statements. One more reason to doubt, figures published on earnings. How many companies re-state their financials? So, at a minimum, inaccurate or maybe, even fraudulent?

 

[MKTrader]

It's not an all or none question. Measuring things like

• Whether a stock index is above its 200-day MA or other long-term trend indicator
• Whether the market has fallen over 20% from an all-time high (classic bear market definition)

are helpful, albeit imperfect measures of the market's current trend. Saying that TA gives you an edge in trading is a very different claim, which is largely untrue with a few exceptions. In fact, defining the long-term trend is one of the few edges if coupled with a profitable strategy (which may or may not involve classic TA).

 

[KCalhoun]

I booked profits today using TA: s&p bearish cup, descending triangle so I went long TZA UVXY

 

[Snuskpelle]

I suppose you could wriggle out of it by saying it is a bull market right up to now, but might go anywhere from now on, but this doesn’t seem to be the way it is used in practice, where bull markets may be deemed to go on for months or years.

The past is a flat fact (probability 1) vs the future which has a range of potential outcomes (w/ a probability distribution). Similarly, you can therefore talk about a piece of the past as if it was a bull or bear market as a fact, but you can't talk about the future like that.

At its simplest, surely TA seeks to find indications that price is a shade more likely to move up or down based on some previous price movements. Isn’t declaring a market to be a bull or bear market implicitly assuming that this is possible and therefore that TA can work? That is, that if the market is moving up it is more likely to continue to do so than not?

The confusion stems from semantics. When people say "TA doesn't work", they generally mean that most of it lacks the predictive qualities many of its proponents claim (and this is easy to confirm by proper back testing). TA works absolutely fine for hindsight analysis, e.g. for figuring out which stocks went up a lot recently. The tool/hammer simile as brought up is generally unhelpful, since people show with it how applicable TA is for hindsight analysis, which nobody was questioning in the first place.

 

[zdreg]

A bull market is when you go consistently short and are losing on nearly every trade.

A bear market is when you go consistently long and are losing on nearly every trade.

 

[deaddog]

The tool/hammer simile as brought up is generally unhelpful, since people show with it how applicable TA is for hindsight analysis, which nobody was questioning in the first place.

Mark Twain said “history never repeats itself but it rhymes"
Trading is a game of probabilities.

 

[MKTrader]

A bull market is when you go consistently short and are losing on every trade.

A bear market is when you go consistently long and are losing on every trade.

Yeah, but if you have some indication of which type of market you're in, that shouldn't happen... you'll still lose some, but certainly not all. Whether you identify the type of market by classic TA measures (above/below 200 day SMA) or by looking at "naked" charts doesn't really matter.

 

[longandshort]

Yeah, but if you have some indication of which type of market you're in, that shouldn't happen... you'll still lose some, but certainly not all. Whether you identify the type of market by classic TA measures (above/below 200 day SMA) or by looking at "naked" charts doesn't really matter.

What type of market we are in only impacts your pnl 1:1 if you’re taking on simple beta. That means you’re not taking active risk.