Yes. they do have to be inefficient for you to capitalize on them. No, they don't. My opinion is they don't because either way prices move. Its just a matter of fact of being on the right side of the trade.
as nononsense seems to be intimating it depends in what way you describe inefficient. An efficient market seems to mean a random market. No markets are completely random or otherwise nobody would trade them. But most markets have a high degree of randomness otherwise everybody would make millions. Actually its a ballancing act, the less randomness in a market the more likely someone will try to exploit the non-random oportunity to profit. When they profit they remove one more opportunity from the market and perhaps make it more random. Of cousre a new competitor could beat an existing competitor if the new trading model is more efficient (at detecting non-randomness).
The market should be inefficient for the system the others use, so that they can lose money; and the market should be efficient for your system so that you can try to catch the money the others lose. But always remember: the others must lose, otherwise you can't win. I think you started from a wrong point of view: not the efficiency of the markets is important, but the efficiency of your system is all that matters. The tradres with the most efficient system will make the money, no matter if the markets are efficient or not.
No, the reason is, prices are not always accurate. To capitalize you just need the price to move, not be inefficient. Inefficiency indicates there being a lag to a change in fundamentals or a loss of something that should be figured/priced in, like some news story that was overlooked, or an eco number thats impact isn't readily factored into the current price: thus a small window of opportunity opens up. While these things probably do happen, and there's plenty of inefficiency in the markets, I've made more money from simple over-reactions or under-reactions allowing me to position myself for the delayed reaction or retracement due to the overaction. So I'd say more profit can be made from these than from inefficiencies. Spike500 is right about the necessity of having an efficient system in order to make money in all kinds of trading environments.
spike500 I'm not sure I agree, I think that a system is effiecient to the degree that it can exploit inefficiencies in the market. Also I must say that I can close my position to you at a profit and you may sell it again at a profit and we both will win. Then the person who buys from you may lose at first but his timeframe may be very long, so he may also win in his timeframe. I'm not saying everyone wins, I'm saying we may each seek to exploit different inefficiencies in the market. How much money we make/lose depends on how true to the market our model is, and the the amount of liquidity which is traded inefficiently in the market.
Toe, are you equating random price moves with inefficiencies? Not disagreeing, I'm just confirming if that is what you mean. Because I never equated randomness with inefficiencies before.
I'm saying randomness in the market is efficiency in the market. The mirror image would be to say that an efficient system is one which exploits non-randomness in the market. If there is only randomness then then the future cannot be reliably forecast. If there is non-randomness then the option is open for someone to exploit it, once non-randomness is fully exploited by traders whats left is whats random, or unexploitable (though I guess many traders trying to exploit one inefficiency may inadvertently create another inefficiency). Mind you I am just a student of quant analisys. The real experts in my sphere are Dr Koch and twiga and others over on WLD web site. I'm using the word "randomness" in a broad sense to include statistical measures like non-stationary, independence, randomness etc. Here's a good resource The Mathematics of Technical Analysis: Applying Statistics to Trading Stocks, Options and Futures by Clifford J. Sherry, Jason W. Sherry