The problem ultimately being that average joe doesnt have a finance degree and wants to get on with their life, and finds it very hard to differentiate between all the different things on offer, peddled by the experts. Thus we end up with the various financial debacles there have been over here in the UK.
Can you predict the future?
- Thread starter zanket
- Start date
zanket:
The stock market is not a zero sum game. The goal of companies is to produce income for their owners.
While I am not 100% sure of what is going on behind the sceens of each trade I am almost positive that I can buy 100 shares and sell 100 shares of Ebay within seconds for a spread of no more than a penny. add that to a commission of $1 and you have a trading cost of $1 + $1 + $0.01 * 100shares/6500 = 3/6500 = %0.02
The stock market is not a zero sum game. The goal of companies is to produce income for their owners.
While I am not 100% sure of what is going on behind the sceens of each trade I am almost positive that I can buy 100 shares and sell 100 shares of Ebay within seconds for a spread of no more than a penny. add that to a commission of $1 and you have a trading cost of $1 + $1 + $0.01 * 100shares/6500 = 3/6500 = %0.02
By zero-sum game, I mean that you can beat the market average only when someone else falls short of it. The average is the return you get for loaning out your money while you hold the stock. It takes no skill to get the average once you’ve bought a diversified portfolio, broad-market index fund or exchange traded fund (like SPY).
Whether you use fundamental analysis or technical analysis, if you try to exceed the market average you are assuming clairvoyance. Some people apparently have this skill, like Warren Buffet. If you don’t then even a cost of 0.02% for switching stock will delay your retirement compared to buy & hold. Also people who predict tend to have fewer stocks in their portfolio; after all, you can’t do good research on 500 stocks. That increases risk which does drag on long-term results. And if you’re American and the account is not tax-deferred (like a 401K), then you’ll get creamed by capital-gains taxes and the paperwork for that. So you should be really sure that you are clairvoyant before trying to beat the market average.
Be wary of those tight Nasdaq spreads. They’ve been wrist-slapped many times over the years for grossly widening the spreads on small orders. A few high-volume stocks do have small spreads, but even institutional investors have a hard time getting 0.1% costs on average for a one-way transaction. I doubt you are really getting 0.02% for a round-trip on EBAY. Consider buying & immediately selling once. If you’re right, then no harm done since the cost was small. If you’re wrong then you’ll have some valuable info.
Whether you use fundamental analysis or technical analysis, if you try to exceed the market average you are assuming clairvoyance. Some people apparently have this skill, like Warren Buffet. If you don’t then even a cost of 0.02% for switching stock will delay your retirement compared to buy & hold. Also people who predict tend to have fewer stocks in their portfolio; after all, you can’t do good research on 500 stocks. That increases risk which does drag on long-term results. And if you’re American and the account is not tax-deferred (like a 401K), then you’ll get creamed by capital-gains taxes and the paperwork for that. So you should be really sure that you are clairvoyant before trying to beat the market average.
Be wary of those tight Nasdaq spreads. They’ve been wrist-slapped many times over the years for grossly widening the spreads on small orders. A few high-volume stocks do have small spreads, but even institutional investors have a hard time getting 0.1% costs on average for a one-way transaction. I doubt you are really getting 0.02% for a round-trip on EBAY. Consider buying & immediately selling once. If you’re right, then no harm done since the cost was small. If you’re wrong then you’ll have some valuable info.
kmguru
Staff member
Just as you do not drive from point A to point B blindfolded, an investment can be managed through business intelligence - ie, if you have enough knowledge about the product and its touch points, then you can predict not far ahead but slightly ahead of most people. Most money managers do not even pay any attention to technology disruptions, how the company is run and basic company information including which idiot suddenly became the sales vice president that can drive the company to the ground.
As a matter of fact, 80% of business analytic systems fail and that is within a company that is supposed to be using it for agility.
As a matter of fact, 80% of business analytic systems fail and that is within a company that is supposed to be using it for agility.
The stock market is one place you can do well blindfolded. Money managers who blindly buy every stock in the S&P 500 (top 500 companies) get the market average. Historically they’ve beaten 80% of the managers who took their blindfolds off and ran up costs by trying to predict future price changes without having the necessary clairvoyance. Really the joke’s on their customers, for the sighted managers often make 10 times more in salary than do the blindfolded managers.
Architectonic
Registered Senior Member
Joeman said:
I know more successful traders who completely ignore most fundamental information for various reasons. (some of which have researched and tested that approach to some extent). One important factor to consider about it, is all solid fundamental information is in the past. Stock prices are based on the future perception of the company - ie the prediction has already been factored in the price.
But indeed, it is true that no crystal ball is required, you do not need to actively predict the future to make money on the market. However, it is a good idea to put everything in your favor. The way to do that is to trade with the trend. If you were to trade with a medium-long term trend (this includes going short in bear markets), you would have averaged far more than 7.5%, with LESS risk, minimal effort. By high and sell higher. Sell low and buy lower.
Zanket - Indeed, liquidity is important to consider - should the stock suddenly move strongly against your position, then you will want to get out as soon as possble, without being screwed by a large spread and overly volatile price movements. High liquidity can sometimes be an indicator that someone knows something about that particular stock.
Last edited:
Trading with a trend is trying to actively predict the future. You are trying to predict how long the trend will last, for which you do need a reliable crystal ball or clairvoyance if the transaction costs will not degrade the performance of your portfolio below that of someone who simply buys & holds. “Buy high and sell higher” assumes that the stock price will go higher after you buy it. But sans predictive ability the odds of that happening are the same for every stock (something like 50.0001% in any given day), so there’s no reason to ever switch stocks.
You can always examine the past to find simple trading strategies that would have significantly beaten the market average. But these strategies have no predictive value; that is, they don’t hold true in the future. For example, I once analyzed several years of winning lottery numbers. The number 9 came up 3 times more often than the least frequent number. But it was literally just luck of the draw. Statistical methods will tell you whether a pattern or strategy is likely just luck or has some merit. There have been very few trading strategies that showed merit, and even those were not profitable after transaction costs.
You can always examine the past to find simple trading strategies that would have significantly beaten the market average. But these strategies have no predictive value; that is, they don’t hold true in the future. For example, I once analyzed several years of winning lottery numbers. The number 9 came up 3 times more often than the least frequent number. But it was literally just luck of the draw. Statistical methods will tell you whether a pattern or strategy is likely just luck or has some merit. There have been very few trading strategies that showed merit, and even those were not profitable after transaction costs.
Last edited:
Architectonic
Registered Senior Member
It is very simple. Due to the nature of the stock market, trending occurs over a time frame. By actively predicting, I mean you consciously decide if the stock is going to go up. If it goes down (significantly, ie not due to noise in the price - time frame is an important factor), you simply sell it without any extra thought and if it goes up you hold it without any extra thought. It does not matter if it magically goes up straight after you sell it. You were not relying on being right more than 50% of the time.
I agree that when you buy a stock, you are making very loose prediction "this may have a good chance of going up". But if it does not, you simply move on. The idea is that you cut your losses sooner than your wins. This way, due to the trending nature of stocks, you will win more than you lose and thus consistently make money. So you rely on the trending nature, rather than actively predicting if a stock will go up or down.
The lottery is completely different - the stock market has a trend bias due to human psychology. If you can't decide if a stock is trending or not, then simply don't buy it.
You will find that these studies that claim that the market does not have any bias, are conducted by close minded economists. Open minded scientists have realized that the markets do have a bias - this bias is human nature.
Last edited:
Architectonic said:
If you cannot predict the future price movements beyond 50% then the transaction costs in selling and re-buying will degrade your performance below that of a buy & hold strategy.
Yes you were. The only way that strategy does better than buy & hold is if you are right more than half of the time.
No, after transaction costs you will lose more than you win this way, unless you are right more than half of the time. Obviously if you are right only half of the time then you will break even before transaction costs regardless of what your trading strategy is.
So you rely on the trend continuing in the future, rather than predicting a price movement? A trend is a price movement.
Most Wall Street pros are believers in the almost pure randomness of the financial markets. They can’t let their customers believe that or else their incomes will plummet. So instead they perpetuate a myth. Ask yourself, if trends are predictable then why don’t the brokerages bet on them rather than just be the house for the bettors? There is a bias but it’s so slight that a retail customer (having relatively large transaction costs) has no hope of exploiting it. Some firms have become adept at exploiting what bias there is; it’s called statistical arbitrage. You need a big bank of computers, a cadre of math whizzes, virtually instant access to the markets, and lots of buying power to keep costs down. These people hope to be right 50.1% of the time. They make their money from the 0.1%.
Last edited:
kmguru
Staff member
Are you saying, human beings are highly unpredictable and therefore the buy and sell reflects that randomness? OR can we be able to capture the 50,000 variables in the decision process and plug into a computer that can track the Buy and Sell habits?
I wonder if the stock market is really a random event like a dice or has a pattern similar to weather which has certain predictability where variables move in a pattern.
I wonder if the stock market is really a random event like a dice or has a pattern similar to weather which has certain predictability where variables move in a pattern.
Future price changes are so unpredictable because market participants are so predictable. It is a very safe bet that their quest for profit will immediately soak up any predictive opportunity in a stock; for example after a takeover announcement. The 50,000 variables are reflected in the price within minutes, not days or weeks. Afterwards there is no information left to make a viable prediction. Those engaged in statistical arbitrage attempt to beat their competitors to the punch by milliseconds.
The stock market is considered to be fractal in nature. As in major trends having minor trends within them. There are eddies of seeming predictability but they discontinue in the future so even those are random near as anyone can tell (well, except for rare individuals like Peter Lynch and Warren Buffet). Hundreds of millions of dollars have been spent trying--and failing--to find predictability. Here’s a recent article about the subject.
The stock market is considered to be fractal in nature. As in major trends having minor trends within them. There are eddies of seeming predictability but they discontinue in the future so even those are random near as anyone can tell (well, except for rare individuals like Peter Lynch and Warren Buffet). Hundreds of millions of dollars have been spent trying--and failing--to find predictability. Here’s a recent article about the subject.
kmguru
Staff member
Hundreds of millions of dollars have been spent trying--and failing--to find predictability.
Yes, that is true...but in the wrong direction. Predictions need a different science than curve fitting the past. Most of the technologies that have been developed tries to find and approximate corelations or teaching a neural net. Several years ago, when I was designing an expert systems for a refinery, I did some serious research to find a technology akin to stock market prediction. One company even sent me a neural software that tracked the stock market since 1970. It tracked so well that I was ready to play the stock market. The fundamental basis was after the fact. So, I had to do my design the hard way. I think, my design can be scaled up to a more predictable environment like terrorist activities but to get to the stock market level - it will require a dynamic model that is beyond present computation and data aquisation level.
Yes, that is true...but in the wrong direction. Predictions need a different science than curve fitting the past. Most of the technologies that have been developed tries to find and approximate corelations or teaching a neural net. Several years ago, when I was designing an expert systems for a refinery, I did some serious research to find a technology akin to stock market prediction. One company even sent me a neural software that tracked the stock market since 1970. It tracked so well that I was ready to play the stock market. The fundamental basis was after the fact. So, I had to do my design the hard way. I think, my design can be scaled up to a more predictable environment like terrorist activities but to get to the stock market level - it will require a dynamic model that is beyond present computation and data aquisation level.
Good stuff. I worked on stock market prediction software for 5 years. I ruled out neural networks and traditional curve fitting early on, in favor of a combination of genetic algorithm and statistics. The hardest technical part was getting decent analysis speed from a standard PC. Database queries were way too slow; instead I did all data mining in memory. The software succeeds in predicting future price movements well beyond luck, but only to an extent that a brokerage house can profit. The hardest part of all has been getting potential customers to take my calls; they’ve been to hell & back trying to doing this, so they’re very skeptical.
Architectonic
Registered Senior Member
How to make money when you are only right 50% of the time. Here, I bring up the old coin toss game - to play the game costs ~$1.00 but if you win, you get $1.10. Of course the coin has a 50% chance of landing on heads or tails, but since you are winning more money than it costs to play, then you make money over the long run.
zanket said:
Architectonic said:
If you lose $1 half the time and gain only 10 cents the other half, you will quickly lose money. The way the market analogously works: You put $1 in the pot. If you have no clairvoyance then you get either 90.9 cents or $1.10 back with 50/50 odds, before transaction costs and before the average 7.5% gain. Play that game with many dollars simultaneously and with any strategy; you'll spin your wheels in the long run. After transaction costs your losses will compound.
They don’t beat the average gain in the forex market without taking greater risk, unless they're clairvoyant.
Architectonic
Registered Senior Member
LOL, that was a typo and a half I meant to say you win $2.10 - or $1.10 after subtracting the cost to play.
If there was a simple strategy that allowed no way to lose then there’d be billions of billionaires out there, don’t you think? The way the market really works is that, absent clairvoyance, you have an equal chance of getting an x% natural logarithmic gain or loss in the short term ignoring transaction costs. Suppose you got a 10% gain; that is, you got back 110% of the money that you put in (you add 1, the 100% you initially wagered). Put 1.1 into the Windows calculator. Press the ln button (natural logarithm). Press the +/- button to make the value negative. Check the inv (inverse, in this case the inverse of the ln function) checkbox. Press ln. The result is the percentage you had an equal chance of getting back. Subtract 1, the 100% you initially wagered. The result is the % loss that you had an equal chance of getting.
For any 10% gain you have an equal chance of a 9.1% loss. For a 100% gain you have an equal chance of a 50% loss. If you play a market that works this way then, absent clairvoyance, transaction costs are highly likely to degrade your performance below the market average. If you are not clairvoyant then your best bet is to buy many stocks (diversify) and don’t sell until you need the money. By diversifying and minimizing transaction costs you get the yield closest to the market average.
For any 10% gain you have an equal chance of a 9.1% loss. For a 100% gain you have an equal chance of a 50% loss. If you play a market that works this way then, absent clairvoyance, transaction costs are highly likely to degrade your performance below the market average. If you are not clairvoyant then your best bet is to buy many stocks (diversify) and don’t sell until you need the money. By diversifying and minimizing transaction costs you get the yield closest to the market average.