Dollar cost averaging a bad investment strategy?

[Flaherty]

I just finished an interesting read on why the investment strategy of dollar cost averaging (DCA) is a bad idea.

DCA is a strategey in which you invest the same amount in an asset every month regardless of the price of the asset. This way, when the price is low you buy more shares and when the price is high you buy fewer. Over the long term, you end up buying shares at the asset's average cost.

The critique I read is highly critical of this strategy.* It's based on the principle that the change in a stock's price is random (I hope that's not a new revelation to you). "A low current stock price forecasts a low future price. If today's price is low, there is a good reason to buy more (it's cheap) and also a good reason to buy less (it's likely to stay cheap). The two reasons cancel out (p.192)" since we cannot predict whether the price will go up or down.

Here's why it's a bad strategy. Imagine walking into a casino and playing 10 different roulette wheels. You bet $1,000 on the first wheel, $2000 on the second wheel, $3000 on the third wheel, and so on, for a total of $55,000. This is a very risky way to play roulette because 60% of your money is at risk on only 3 turns of the wheel. A lower risk strategy is to bet $5500 on each of the 10 wheels.

DCA is just like playing the casino as described above. Whether your stock goes up or down a given month is just as random an outcome as a spin of the roulette wheel. So, if you buy $1000 worth of shares in January and repeat this each month for the next 10 months, then the market's outcome in August, September, and October will dictate 60% of your annual return.

It would be wiser to invest $5500 the first month and then adjust your holdings up or down each month to keep the value of the holdings constant. If the value of your portfolio goes down $500, then add it back with by buying another $500 in shares. If it goes up $500 then sell the profit to keep the total value at $5500.

Now imagine that the value of the stock goes up in 6 of the next 10 months and down the same amount in the other 4 months. If you keep your holdings constant as described in the previous paragraph, then you will come out ahead guaranteed. But if you DCA, then it matters which months the price goes up and which months it goes down. If the good months are the early ones, then the DCA strategy will be a big loser. DCA adds an unnecessary dimension of risk to your investment.

Are there other viewpoints on DCA?

* From economist Steven Landsburg, The Armchair Economist.

[jimlintott]

Dollar cost averaging is a very good way to invest for many people. First, many people of modest incomes don't have large amounts of money to play with but can afford a certain amount each month. Second, you don't have to overly concern yourself with the markets. Particularly if buying mutual funds. Trying to time the market is a difficult task that few of us have time to do.

I would tend to agree that if you have a large amount of money to invest then you should invest it all and not dole it out piece by piece. Having a large sum in savings won't get you the results you are probably after. (Although it would be a safe investment.)

The stock market isn't random. It may appear that way but it in fact respond to actual events and often quite predictably. If the market was random then prices would just jump around, well, randomly. The truth is that prices respond to real world events like quarterly reports, terrorist attacks, etc. Some of these events may be predictable and some aren't.

[Drooper]

Quote:

Originally posted by Flaherty
I just finished an interesting read on why the investment strategy of dollar cost averaging (DCA) is a bad idea.

.....

The critique I read is highly critical of this strategy.* It's based on the principle that the change in a stock's price is random (I hope that's not a new revelation to you). "A low current stock price forecasts a low future price. If today's price is low, there is a good reason to buy more (it's cheap) and also a good reason to buy less (it's likely to stay cheap). The two reasons cancel out (p.192)" since we cannot predict whether the price will go up or down.


Here's why it's a bad strategy. Imagine walking into a casino and playing 10 different roulette wheels. You bet $1,000 on the first wheel, $2000 on the second wheel, $3000 on the third wheel, and so on, for a total of $55,000. This is a very risky way to play roulette because 60% of your money is at risk on only 3 turns of the wheel. A lower risk strategy is to bet $5500 on each of the 10 wheels.


DCA is just like playing the casino as described above. Whether your stock goes up or down a given month is just as random an outcome as a spin of the roulette wheel. So, if you buy $1000 worth of shares in January and repeat this each month for the next 10 months, then the market's outcome in August, September, and October will dictate 60% of your annual return.

It would be wiser to invest $5500 the first month and then adjust your holdings up or down each month to keep the value of the holdings constant. If the value of your portfolio goes down $500, then add it back with by buying another $500 in shares. If it goes up $500 then sell the profit to keep the total value at $5500.

Now imagine that the value of the stock goes up in 6 of the next 10 months and down the same amount in the other 4 months. If you keep your holdings constant as described in the previous paragraph, then you will come out ahead guaranteed. But if you DCA, then it matters which months the price goes up and which months it goes down. If the good months are the early ones, then the DCA strategy will be a big loser. DCA adds an unnecessary dimension of risk to your investment.

Are there other viewpoints on DCA?

* From economist Steven Landsburg, The Armchair Economist.

I think this is all wrong.

Starting from the assumption that the stock market is efficient (all information is priced), then making any investment decision based on price movement (as you advocate: "If the value of your portfolio goes down $500, then add it back with by buying another $500 in shares") means you are taking on additional risk for no adidtional expected return.


Also, you contradict youself.

Buying$5,000 in one tranche will include more risk than buying in 5 tranches of $1,000, but offer the same expected return.Think again about your Casino example; you are correct, but seem to get the moral confused when you apply it to stocks.



It is the case that if you have an investment plan where you invest a certain sum each month, you will sort of iron out the unpredictable monthly movements in the share price.

If you just make one lump purchase, you open youself up to higher risk, because you may have bought when the price just happened to be high. (of course you may have bought when it happened to be low - but you won't know that at the time).

[Walter Wayne]

Quote:

Originally posted by Flaherty
It would be wiser to invest $5500 the first month and then adjust your holdings up or down each month to keep the value of the holdings constant. If the value of your portfolio goes down $500, then add it back with by buying another $500 in shares. If it goes up $500 then sell the profit to keep the total value at $5500.

Note also that you don't invest all your money in this system. You need a larger seed than your initial investment. For example if you lose 10% in the first quarter, you need to come up with $550. In addition you will never see the magic of compounding with this.

The other thing they misrepresent here, is calculating the gain over a single year. Most low to mid income people are saving for the long term. With the standard DCA you have done most of your investing by the time you are 40 or 50, at which point even moderate interest swamps your contribution. Our fiscal cycle isn't a year it is a lifetime (invest in education and job, invest money while we are at a job, retreive investments when retiring).

I would think (based on math) DCA is better long-term with small initial investment.

Walt

[Iconoclast]

Quote:

Originally posted by Flaherty
The critique I read is highly critical of this strategy.* It's based on the principle that the change in a stock's price is random (I hope that's not a new revelation to you).

For those who invest based solely on stock price charts, the changes in the price of a stock could be considered random, however the reality is that these price fluctuations are based on real forces.

Quote:

Originally posted by Flaherty
"A low current stock price forecasts a low future price. If today's price is low, there is a good reason to buy more (it's cheap) and also a good reason to buy less (it's likely to stay cheap). The two reasons cancel out (p.192)" since we cannot predict whether the price will go up or down.

The two reasons cancel out? I'm sorry, but I don't think that's a mathematically valid assertion.

Quote:

Originally posted by Flaherty
Here's why it's a bad strategy. Imagine walking into a casino and playing 10 different roulette wheels. You bet $1,000 on the first wheel, $2000 on the second wheel, $3000 on the third wheel, and so on, for a total of $55,000. This is a very risky way to play roulette because 60% of your money is at risk on only 3 turns of the wheel. A lower risk strategy is to bet $5500 on each of the 10 wheels.

Well, in both scenarios the expected value is identical, and in the first example the variance will be very much higher. Is that what you meant to say? Your second example may be considered more "risk averse", but only so far as your bankroll will experience fluctuations of a smaller magnitude.

Quote:

Originally posted by Flaherty
DCA is just like playing the casino as described above. Whether your stock goes up or down a given month is just as random an outcome as a spin of the roulette wheel. So, if you buy $1000 worth of shares in January and repeat this each month for the next 10 months, then the market's outcome in August, September, and October will dictate 60% of your annual return.

It would be wiser to invest $5500 the first month and then adjust your holdings up or down each month to keep the value of the holdings constant. If the value of your portfolio goes down $500, then add it back with by buying another $500 in shares. If it goes up $500 then sell the profit to keep the total value at $5500.

Surely a much better strategy would be to attempt to analyse the company, any recent management decisions that may (in your opinion) make the stock more or less desirable, any changes to regulatory laws in the area of business in which this company trades, any natural or political events that may affect how this stock is viewed. To sell $500 worth of shares in a company that is experiencing boom times just so that you can keep your exposure to some magical $5000 value is stupidity pure and simple. Adding another $500 to your investment in that same stock only because it lost value is a similarly poor decision.

[ebola]

Flaherty,

Take this book back to the retailer and demand your money back. This guy is an idiot who should not be dispensing financial advice. He is obviously more comfortable doing his investing in large, windowless buildings which echo the sounds of slot machines.

Quote:

Flaherty wrote:

DCA is a strategey in which you invest the same amount in an asset every month regardless of the price of the asset. This way, when the price is low you buy more shares and when the price is high you buy fewer. Over the long term, you end up buying shares at the asset's average cost.

I assume you are paraphrasing him here. The part in bold is misleading. For example, if you choose to start DCA $100/month in a given stock, and the first month the price is $40, and the second month the price is $25, the average cost is $37.50, but you own 6.5 shares which cost you $200. Your cost basis per share is $30.78, which is considerably less than the average cost. This is one of the strengths of DCA.

DCA allows you to accumulate shares of stock ( or mutual funds ) over time, and your cost basis ( neglecting fees ) will always be below the arithmatic mean of the purchase prices. It focuses on a long term, buy and hold strategy which is the only way the average investor will come out on top in the stock market.

Quote:

The critique I read is highly critical of this strategy.* It's based on the principle that the change in a stock's price is random (I hope that's not a new revelation to you). "A low current stock price forecasts a low future price. If today's price is low, there is a good reason to buy more (it's cheap) and also a good reason to buy less (it's likely to stay cheap). The two reasons cancel out (p.192)" since we cannot predict whether the price will go up or down.

I disagree here too. The stock's prices, while they may appear to be random, are not. No investor should be accumulating a given asset without doing some form of research to verify a solid corporate foundation. Also, a current low price does not forecast a future low price. How many people have you heard a cocktail parties talk about how they ( or a family member ) had the opportunity to buy Microsoft ( or other big winner ) at an unbelievably low, split adjusted price just a few short years ago.

While there are examples of low current prices forecasting low future prices ( Enron, WorldCom, et al. ), these are lessons for diversification, an extremely important concept in any investment strategy. If you had all of your money in Enron, you lost your shirt. If you were diversified over 20 stocks, you took a hit, but you didn't lose everything, unless you chose very, very badly.

One other thing; to make money in the stock market, everybody knows that you buy low and sell high. I don't think that anybody has a problem with the sell high part; the difficulty arises when presented with the chance to buy low. This brings us to the problem of market timing:

Quote:

jimlintott wrote:

Trying to time the market is a difficult task that few of us have time to do.

I disagree. Market timing is an impossible task that the most highly skilled professionals CANNOT do. The only way that anyone can be reasonably sure to buy low and sell high is to buy solid companies, hold their stock for years, and then sell. Warren Buffet invests this way. In addition, he isn't afraid to invest in a stock if its price has been beaten down. In fact, for people who invest like he does, low stock prices represent an opportunity.

Eric

[ebola]

There is one thing I forgot to mention:

While stock prices do fluctuate up and down over the short term, they do trend upwards over the long term. This is why long term buy and hold works: If you were to buy SPY ( S&P depository receipts; an exchange traded fund which mirrors the S&P 500 ) at the peak price this year, and 20 years from now sell them at their lowest point that year, you would likely have a return that would mirror the historical returns of the stock market, despite the fact that your timing stank. This is of course, not guaranteed, as it is true that ( as it says on the cover of every mutual fund prospectus ) past performance is no guarantee of future results. However, this has been consistently true for over eighty years.

Eric

[Iconoclast]

Quote:

Originally posted by Flaherty
* From economist Steven Landsburg, The Armchair Economist.

Flaherty, your first post wasn't quite clear. Is the book your read by Steven Landsburg, or was the critique by him?

I ask because a quick scan of the reader reviews for "The Armchair Economist" seems to indicate that in the last chapter of his book he talks about a scathing letter he sent to his daughter's kindergarten teacher in response to her teaching the value of recycling. According to him (again this is from the reader review) recycling is bad for the environment because paper producers plant less new trees when paper recycling occurs. If his views on economics are as sound as his understanding of the environment, then as far as I'm concerned Mr Landsburg can stick his theories right up his bottom.

[Flaherty]

Quote:

Originally posted by jimlintott
The stock market isn't random. It may appear that way but it in fact respond to actual events and often quite predictably. If the market was random then prices would just jump around, well, randomly. The truth is that prices respond to real world events like quarterly reports, terrorist attacks, etc. Some of these events may be predictable and some aren't.

Oh, I agree that a stock's price depends greatly on its expected earnings, interest rates, and the like. However, the point is that changes in the stock's price are random because future events pop up randomly.

For example, we all agree that if the firm's next earnings report comes in above expectations that the price will almost certainly go up. The trick is that we cannot guess better than 50/50 whether the next report will be favorable or unfavorable. This is the random element.

Imagine a roulette wheel where each number represents a kind of news event that may affect the stock's price that day. Maybe there are 100 spaces on the wheel. If it's a good company, than most of the spaces will be events favorable to the company's value, like "sales rise," or "productivity increases." But some of the spaces will be bad things like "earnings below expectations," or "labor strike," or "interest rates rise," or "board fires CEO, hires a 10-year old." Although we have a good idea what will happen to the stock price if any particular event pops up, we have no way of knowing which event will pop up. That is the random nature of the market.

[Flaherty]

Quote:

Originally posted by Iconoclast
Flaherty, your first post wasn't quite clear. Is the book your read by Steven Landsburg, or was the critique by him?

I ask because a quick scan of the reader reviews for "The Armchair Economist" seems to indicate that in the last chapter of his book he talks about a scathing letter he sent to his daughter's kindergarten teacher in response to her teaching the value of recycling. According to him (again this is from the reader review) recycling is bad for the environment because paper producers plant less new trees when paper recycling occurs. If his views on economics are as sound as his understanding of the environment, then as far as I'm concerned Mr Landsburg can stick his theories right up his bottom.

Sorry for the confusion. The book is by Landsburg and in it he critiques DCA.

The chapter on environmentalism is certainly controversial, but I think he has the paper recycling aspect of it 100% correct.

Producing virgin paper requires that trees be cut down and processed. However, paper companies plant new trees to harvest them every 30 years or so just for this purpose. If more people use virgin paper, then more trees will be cut down and more trees will be planted.

The environmentalists' error is assuming that using virgin paper necessarily results in there being fewer trees. This might very well be the case, but it is not automatically the case because market incentives work in both directions. The question has to be settled by empirics.

People eat lots of steak, yet there is no shortage of cattle. This is because by purchasing a steak, we send a signal to producers to make more cattle. Likewise, when we buy paper we send a signal to producers to make more trees.

Landsburg's larger point in that chapter is a good one that any skeptic should appreciate. He notes there is a lot of emotionalism when it comes to environmentalism and that this gets in the way of clear, reasoned thinking.

For example, it is common to teach school children that the Amazon rainforest is shrinking x% per year. This x% number is meaningless unless illuminated by theory. Is the x% shrink rate too fast, too slow, or just right? There are indeed benefits to cutting down forests: wood for paper, wood for homes for humans, new farmland, and so on. There are also costs: destruction of natural habitat for many species, lost beauty of a rain forest, climactic effects, and so on.

What Landsburg (and I) are saying is that the benefits have to be weighed against the costs. In other words, the question should be settled by empirics rather than emotionally assuming that it is always good to preserve the rainforest. If the net benefits are high enough, then maybe the rainforest is not shrinking fast enough! The point is that we don't automatically know whether the net benefit is positive or negative, so it is rash to make an assumption in that regard.

[Flaherty]

Quote:

Originally posted by Walter Wayne
Note also that you don't invest all your money in this system. You need a larger seed than your initial investment. For example if you lose 10% in the first quarter, you need to come up with $550. In addition you will never see the magic of compounding with this.

No, I don't think that's what it is saying.

You can invest all or part of your money; just diversify, which is what you're supposed to do anyway.

Using the numbers from my original post, if your stock value goes up from $5500 to $6000 after one month, then sell $500 worth and put it into another asset. The magical compounding is still there, it's just diversified into multiple assets.

If events convince you the underlying company is better (worse) than you initially thought, then revise your target holding amount to something higher (lower).

[Iconoclast]

Quote:

Originally posted by Flaherty
The chapter on environmentalism is certainly controversial, but I think he has the paper recycling aspect of it 100% correct.

Producing virgin paper requires that trees be cut down and processed. However, paper companies plant new trees to harvest them every 30 years or so just for this purpose. If more people use virgin paper, then more trees will be cut down and more trees will be planted.

The environmentalists' error is assuming that using virgin paper necessarily results in there being fewer trees.

I don't think so. Using more virgin paper and not recycling the old stuff necessarily results in larger timber plantations, and this is a bad thing. Recycling will result in a reduction in the amount of plantation timber required annually and hence a reduction in the total size of plantations. Remember that plantations cannot support wildlife in the way old growth forrests can since there is no accumulation of undergrowth, they're nothing more than tree farms. Then there's the massive amount of extra land fill created because we've chosen to bury all our old paper and cardboard instead of recycling it all. Then there's the extra stress on the soil because plantation timber isn't allowed to die and rot which returns to the soil all of the nutrients the tree extracted during it's lifetime, so fertiliser must be used to grow the new trees.

Quote:

Originally posted by Flaherty
People eat lots of steak, yet there is no shortage of cattle. This is because by purchasing a steak, we send a signal to producers to make more cattle.

We don't recycle steaks because we consume it, it's food for us like grass was food for the cows.

Quote:

Originally posted by Flaherty
Landsburg's larger point in that chapter is a good one that any skeptic should appreciate. He notes there is a lot of emotionalism when it comes to environmentalism and that this gets in the way of clear, reasoned thinking.

I don't think the author has thought enough about that issue to make a reasoned decision.

Quote:

Originally posted by Flaherty
If the net benefits are high enough, then maybe the rainforest is not shrinking fast enough!

That's a pretty arrogant fecking assumption when you consider just how much wilderness has been lost in the last hundred years.

[Flaherty]

Quote:

Originally posted by Iconoclast

I don't think so. Using more virgin paper and not recycling the old stuff necessarily results in larger timber plantations, and this is a bad thing. Recycling will result in a reduction in the amount of plantation timber required annually and hence a reduction in the total size of plantations. Remember that plantations cannot support wildlife in the way old growth forrests can since there is no accumulation of undergrowth, they're nothing more than tree farms. Then there's the massive amount of extra land fill created because we've chosen to bury all our old paper and cardboard instead of recycling it all. Then there's the extra stress on the soil because plantation timber isn't allowed to die and rot which returns to the soil all of the nutrients the tree extracted during it's lifetime, so fertiliser must be used to grow the new trees.

Some of these are legitimate costs of using virgin paper and should be factored into any good cost-benefit study of the effects of recycling. However, recycling is not free, either. My city had to buy lots of recycling trucks, pay workers to go around collecting materials, and build a recycling plant to process all this stuff.

Don't get me wrong here; recycling paper may indeed be a net plus for society. However, it's wrong for people to assume that because some action protects the environment that it is necessarily a good idea. We can have too much environmental protection. The real trick is finding the right amount; the amount of conservation that balances the benefits with the costs.

Quote:

We don't recycle steaks because we consume it, it's food for us like grass was food for the cows.

The principle is the same. Not recycling paper does not necessarily result in fewer trees being around in the same way that not recycling steak does not result in fewer cattle.

Quote:

That's a pretty arrogant fecking assumption when you consider just how much wilderness has been lost in the last hundred years.

No one questions whether many acres of wilderness have been lost. The question is whether the other uses for which those acres are being used is greater than the cost of losing the forests that once stood there. What is the net benefit or loss? This is not arrogance; it's a rational question. Environmentalism doesn't get a free pass from critical, skeptical analysis.

[phildonnia]

Ok, suppose a stock is $10 half of the time, and $20 half of the time. And a total of $200 is invested.

With cost-averaging, $100 is spent on the $10 stock, and $100 is spent on the $20 stock. A total of 15 shares are purchased.

If you buy it all at once, you get either 20 shares or 10 shares, depending on the price. On average, you get 15 shares.

So I see no real difference in the expected return, but there's more certainty with the cost-averaging. Am I missing some subtlety here known only to financial advisors?

[phildonnia]

Originally posted by Flaherty
If the net benefits are high enough, then maybe the rainforest is not shrinking fast enough!

Originally posted by Iconoclast
That's a pretty arrogant fecking assumption when you consider just how much wilderness has been lost in the last hundred years.

Not to join the hijackers, but I could not find an assumption, fecking or otherwise, in Flaherty's hypothetical.

[Iconoclast]

Quote:

Originally posted by phildonnia
Not to join the hijackers, but I could not find an assumption, fecking or otherwise, in Flaherty's hypothetical.

Fine, assumption was the wrong word, but personally I can think of NO scenario that would cause me to exclaim "The amazon isn't disappearing fast enough".

[Number Six]

Quote:

Originally posted by Flaherty
I just finished an interesting read on why the investment strategy of dollar cost averaging (DCA) is a bad idea.

DCA is a strategey in which you invest the same amount in an asset every month regardless of the price of the asset. This way, when the price is low you buy more shares and when the price is high you buy fewer. Over the long term, you end up buying shares at the asset's average cost.

The critique I read is highly critical of this strategy.* It's based on the principle that the change in a stock's price is random (I hope that's not a new revelation to you). "A low current stock price forecasts a low future price. If today's price is low, there is a good reason to buy more (it's cheap) and also a good reason to buy less (it's likely to stay cheap). The two reasons cancel out (p.192)" since we cannot predict whether the price will go up or down.

Here's why it's a bad strategy. Imagine walking into a casino and playing 10 different roulette wheels. You bet $1,000 on the first wheel, $2000 on the second wheel, $3000 on the third wheel, and so on, for a total of $55,000. This is a very risky way to play roulette because 60% of your money is at risk on only 3 turns of the wheel. A lower risk strategy is to bet $5500 on each of the 10 wheels.

DCA is just like playing the casino as described above. Whether your stock goes up or down a given month is just as random an outcome as a spin of the roulette wheel. So, if you buy $1000 worth of shares in January and repeat this each month for the next 10 months, then the market's outcome in August, September, and October will dictate 60% of your annual return.

It would be wiser to invest $5500 the first month and then adjust your holdings up or down each month to keep the value of the holdings constant. If the value of your portfolio goes down $500, then add it back with by buying another $500 in shares. If it goes up $500 then sell the profit to keep the total value at $5500.

Now imagine that the value of the stock goes up in 6 of the next 10 months and down the same amount in the other 4 months. If you keep your holdings constant as described in the previous paragraph, then you will come out ahead guaranteed. But if you DCA, then it matters which months the price goes up and which months it goes down. If the good months are the early ones, then the DCA strategy will be a big loser. DCA adds an unnecessary dimension of risk to your investment.

Are there other viewpoints on DCA?

* From economist Steven Landsburg, The Armchair Economist.

I think this person's analysis is wrong. First of all, there are two possible scenarios, those being

1) You have a chunk of money to invest right now

2) You get a certain amount of money each month to invest.

If Case #2 holds then by definition of the market being random (whcih is isn't strictlyspeaking, but which it is effectively speaking for most investors) DCA holds. You choose your favorite allocation of stocks, bonds, etc, (the best allocation being random as far as you can tell, therefore you have no idea if any allocation you choose is good or bad) and then you invest your dollar amount in that allocation each month.

If Case #1 holds then you have a big chunk of money to invest right now (say, $55,000). The author makes an analogy to investing $1,000 on roulette wheel #1, $2,000 on roulette wheel #2, etc, up to $10,000 roulette wheel #10. I don't think this is analogous at all. Why is that DCA? DCA would be investing $5,500 each month for 10 months, while temporarily holding the uninvested money in some low risk, low return account. It's possible that you'll end up buying low, in which case you'll make out. Or it's possible you'll end up buying high, in which case you'll lose out.

But the point is, if the market truly is RANDOM (as far as you can tell at least) then YOU DON'T KNOW WHAT WILL HAPPEN and the chance of winning and losing are the same and by assuming randomness you're expected return is the SAME while you simulatneously mimimize your maximium loss (or maximum gain).

If you have a random process then DCA will minimize the variation in your investment. If you want to take a chance, possibly winning big and possibly losing big, then put the whole wad on a single bet. And that may make sense if the market will not go up over time. After all, if the market is going to go up and down for the next 20 years but you know that 20 years from now it's going to be the same place that it is now, then you might as well put all your money in it now and hope that a year from now it is up, in which case you can take your profits and run.

But if you think the market going to go up in the long run then DCA makes sense because it says that you'll be nearer the average, and that average will be up.

In fact, that is the bottom line...DCA means there is a greater chance you'll be near the average and not doing DCA means there is a greater chance you'll not be near the average. If the average goes up over time (as it has done with the stock market) then you gain over time with DCA whereas with non-DCA then you might win, and maybe even win bigger than DCA wins, but you also may lose.

If the market is random (as near as you can tell) then the expected return is ALWAYS the same regardless of trading strategy (trading costs, etc, aside). DCA just minimizes variation.

[TeaBag420]

These extended quotations of previous posts have become tedious.

Marketing timing can be and has been done successfully:


http://www.bobbrinker.com

[Drooper]

Quote:

Originally posted by Iconoclast
Fine, assumption was the wrong word, but personally I can think of NO scenario that would cause me to exclaim "The amazon isn't disappearing fast enough".

No scenario?

What about this scenario.

Braziil enacts a law stopping any future land clearance in the country.

However, Brazil is extremely poor (or a large proportion of the population, especially in rural areas is). This transaltes into extremely high infant mortality, diseases that have been eradicated in the west, little education, poor health system etc.

Some clearance of the Amazon, would bring economic acitivity, income and improve wealfare, at the expense of the loss of a tiny proportion of the total forested area in Brazil. Could you not "think" that this tradeoff might be acceptable in the scenario?

Could you not think that in this case, the Amazon would not be "disappearing" fast enough?


An extreme example, maybe, but you need to accept the principle. In wealthy countries, we simply take it for granted that one of the reasons why 98% of children survive until the age of 4, rather than 50% is partly down to the fact that our anscestors pillaged the natural resources of the land, including massive (in the case of the UK almost total) deforestation.

[Drooper]

Quote:

Originally posted by Flaherty
I just finished an interesting read on why the investment strategy of dollar cost averaging (DCA) is a bad idea.

Flaherty,

I have just been thinking about this and decided to modify my previous response.

After Googling DCA, I find that I am of the opinion that this is a sensible way of investing over time. However, the stated benefits of DCA are false.

By this I mean that the point that many here have made, including me, is that investing small sums over time has zero effect on expected return, but will reduce the risk (standard deviation, or variance) of the return. For this reason it can make sense.

However, the claimed benefit; that DCA leads you to buy more stock at a cheaper price is bunk, for the reasons pointed out by your "Armchair Economist". Namely, the market is efficient.

[Ladewig]

On topic: ixnay on criticizing investment techniques for small investors - we're trying to start another bubble.

Off topic: The unstated assumption on the its-not-shrinking-fast-enough scenario is that if after a period of time the negative effects outweigh the positive effects, we can simply replant the rainforest to reduce the negative effects.

[Drooper]

Quote:

Originally posted by Ladewig
Off topic: The unstated assumption on the its-not-shrinking-fast-enough scenario is that if after a period of time the negative effects outweigh the positive effects, we can simply replant the rainforest to reduce the negative effects.

No, there is no such assumption. It is a question of evaluating whether the marginal benefits of clearance exceed the costs. No assumptions.

[ebola]

Quote:

TeaBag420 wrote:

Marketing timing can be and has been done successfully:


http://www.bobbrinker.com

1) What is his net worth ?

2) How much of it came from trading, and how much is a result of him peddling his wares ?

3) Is his net worth even close to Warren Buffet's ?

4) If his strategy is so effective, why is he telling us ? Wouldn't he be content to crush the market, year after year ? Wouldn't this maximize his wealth ? Or is he selling financial advice because it brings in a higher and more certain income than his " sure thing " ?

Some market timers have been successful for short periods of time. Predicting the market accurately over an extended period of time is a fool's errand.

Eric

[Iconoclast]

Quote:

Originally posted by Drooper
No scenario?

Well, maybe a completely contrived scenario could change my mind...

Quote:

Originally posted by Drooper
What about this scenario.

Braziil enacts a law stopping any future land clearance in the country.

Some clearance of the Amazon, would bring economic acitivity, income and improve wealfare, at the expense of the loss of a tiny proportion of the total forested area in Brazil. Could you not "think" that this tradeoff might be acceptable in the scenario?

So, you want to know if Brazil completely stopped land clearing, then later on allowed a teeny tiny little bit of clearing, could it be said that the Amazon is disappearing at too slow a rate? Well in that completely contrived scenario I just might say that. Sorry Drooper, but my post was about the reality of land clearing as it stands today. From the People and Planet web site (Feb 2001):

"New research suggests that up to 95 per cent of the Amazon rainforest could be destroyed over the next 20 years, compared with just 4 per cent between the 'discovery' of Brazil in 1500 and the end of the 1970s. The speed of destruction doubled over the past 20 years, clearing an area the size of France at the rate of 20,000 sq km a year."

I really don't think we should accelerate the rate of land clearing, do you?

Well I think this thread has now been well and truly hijacked -- and it's all my fault -- but getting back to what Mr Lansburg stated, he said that paper recycling was bad for the environment because more trees have to be planted if paper isn't recycled. His primary error is in assuming that all forests are created equal. From an environmental point of view, 100 square miles of old growth rainforest is vastly more valuable than 100 square miles of timber plantation, so we should be trying to minimise the area required for plantation timber by recycling paper. I really didn't think this point was rocket science.

[Iconoclast]

Quote:

Originally posted by Ladewig
Off topic: The unstated assumption on the its-not-shrinking-fast-enough scenario is that if after a period of time the negative effects outweigh the positive effects, we can simply replant the rainforest to reduce the negative effects.

And that assumption is a poor one. If virgin forest is cleared and then years later we realise we've made a mistake, we can't just "replant the rainforests". You're talking about an ecosystem that to be brought back to it's original state would take something in the order of thousands of years. You can't just replant a rainforest and come back in twenty years and expect it to be back to the same state it was before it was cleared.

[Ziggurat]

Quote:

Originally posted by Drooper

By this I mean that the point that many here have made, including me, is that investing small sums over time has zero effect on expected return, but will reduce the risk (standard deviation, or variance) of the return. For this reason it can make sense.

However, the claimed benefit; that DCA leads you to buy more stock at a cheaper price is bunk, for the reasons pointed out by your "Armchair Economist". Namely, the market is efficient.

I think one reason you're missing the return advantage of DCA is you're not actually considering what the alternative really is, namely share cost averaging. Investing all your money up front is pretty much never an option, unless you're a trust fund baby. For simplicity, let's consider a stock that sells for $1 on one date and $2 on another date. Let's say you DCA $20 into this sock, buying 10 shares one one date and 5 shares on a second date, for a total of 15 shares and $1.33/share average cost. Now let's instead suppose you share cost averaged, and bought 7.5 shares on the first date (cost $7.5) and 7.5 shares on the second date (cost $15), for a total of 15 shares (same final number), but at a total cost of $22.50 and an average cost of $1.50.

Yes, the primary benefit of DCA is reduction of risk, but the return difference compared to the alternative is real. When buying stock, you should dollar cost average. When selling stock, you should share cost average.

[Ladewig]

Quote:

No, there is no such assumption. It is a question of evaluating whether the marginal benefits of clearance exceed the costs. No assumptions.

Will you admit that there is an assumption that we can identify and quantify the long-term benefits and costs of clear cutting the forest? I question the validity of that assumption.

[TeaBag420]

Quote:

Originally posted by ebola


1) What is his net worth ?

Don't know, but I know what mine is.

Quote:

2) How much of it came from trading, and how much is a result of him peddling his wares ?

When you've read the website I would be happy to discuss that with you.

Quote:

3) Is his net worth even close to Warren Buffet's ?

You must be referring to the man Bob describes as "the Sage of Omaha". By the way, do you know Mr. Buffett's annual salary? I DO. It's $100,000. I do sometimes research things, as sceptics are wont to do.

Quote:

4) If his strategy is so effective, why is he telling us ?

Let me turn that around: if you know what you are talking about, why are you telling us. Game, set, match. Reduction ad absurdum. TDLA.

Geez, I dunno, to help people, to have something to do, to have a reason to get up in the morning? Maybe he just likes having a radio show. Where can I hear your radio show?

Quote:

Wouldn't he be content to crush the market, year after year ? Wouldn't this maximize his wealth ?

Why must you try to stuff your peni... uh, words into Mr. Brinker's mouth. Can you find one use of the phrase "sure thing" on either his website or his archived radio shows?

You don't need to "crush" (your word) the market to generate wealth. Beating it suffices quite nicely. Matching it works too.

Quote:

Or is he selling financial advice because it brings in a higher and more certain income than his " sure thing " ?

It's called "keeping your day job"... it's part of "diversification" <---- that's a financial term.

Quote:

Some market timers have been successful for short periods of time. Predicting the market accurately over an extended period of time is a fool's errand.

Eric

And you present what facts and figures in support of that argument by assertion?


Would you consider sixteen years an extended period of time?

[The Central Scrutinizer]

Quote:

Originally posted by TeaBag420
These extended quotations of previous posts have become tedious.

Marketing timing can be and has been done successfully:


http://www.bobbrinker.com

Or so he claims.


Funny that he happens to sell stuff too.

[TeaBag420]

Quote:

Originally posted by The Central Scrutinizer


Or so he claims.


Funny that he happens to sell stuff too.

Yeah, trudat homie! Of course Warren Buffet, the man Bob calls the Sage of Omaha, SELLS SHARES IN BERKSHIRE HATHAWAY. Pyramid scheme? (begin Merv Griffin voice) Oooooh! (end Merv Griffin voice).

And don't try to step to me with Zappa. I'll eat your snouts and your trotters first.

[Drooper]

Quote:

Originally posted by Ladewig


Will you admit that there is an assumption that we can identify and quantify the long-term benefits and costs of clear cutting the forest? I question the validity of that assumption.

So you mean that we cannot assume that the rainforest has any value at all?

So we just chop it all down?

No, I think it has value and I know we have methods of quantifying such value.

[Ladewig]

Quote:

So you mean that we cannot assume that the rainforest has any value at all?

So we just chop it all down?

No, I think it has value and I know we have methods of quantifying such value.

My point is that any cost/benefit analysis related to the rate of cutting down the Amazonian rainforest should include both short-term costs and benefits and long-term costs and benefits. I maintain that the long-term costs and benefits might be correctly identified, but cannot be accurately quantified. People willing to perform such a cost/benefit analysis of cutting down trees are either assuming that they can quantify these long-term factors or are assuming that the long-term factors can be ignored. I question the validity of that assumption.

[Drooper]

Quote:

Originally posted by Ziggurat


I think one reason you're missing the return advantage of DCA is you're not actually considering what the alternative really is, namely share cost averaging. Investing all your money up front is pretty much never an option, unless you're a trust fund baby. For simplicity, let's consider a stock that sells for $1 on one date and $2 on another date. Let's say you DCA $20 into this sock, buying 10 shares one one date and 5 shares on a second date, for a total of 15 shares and $1.33/share average cost. Now let's instead suppose you share cost averaged, and bought 7.5 shares on the first date (cost $7.5) and 7.5 shares on the second date (cost $15), for a total of 15 shares (same final number), but at a total cost of $22.50 and an average cost of $1.50.

Yes, the primary benefit of DCA is reduction of risk, but the return difference compared to the alternative is real. When buying stock, you should dollar cost average. When selling stock, you should share cost average.

By using this "DCA" all you are doing is allowing the number of shares purchased with each tranche to vary.

The only way that could lead to a higher expected return, compared with buying a constant number of shares is if a lower share price was associated with a higher expected return.

If the expected return at $1 is, say 5% and the expected return at $2 is also 5%, how can different combinations of purchases at the different prices alter the expected return?

You are using this DCA as an implicit contra-indicator of expected return, by saying that by buying more shares when the price goes down increases your expected return, which is wrong unless you have some information that the market is not pricing.

[Number Six]

As far as the website with the guy telling you how to get rich by market timing, you don't have to read it to know it's very likely to be BS. This is like standing in line at the grocery store and seeing the little booklets for say for 49 cents that have lucky lottery numbers. If you can beat the lottery then you play the lottery, not try to sell numbers to other people. Similarly, if you can time the market you time the market, not try to make money off of telling other people how to time the market.

Keeping a radio show on the side isn't what they mean when they say diversification. Rather that means to diversity _within_ the market. If you could time the market you wouldn't need a radio show outside it. For that matter, if you coudl time the market you wouldn't even need diversification _within_ the market. All you'd need would be one aspect of the market that you could consistently time.

And it can't be that that guy just wants to spread happiness by having more people use his method and getting everyone rich because by definition the more people that use any method the less effective that method is.

And what relevance does Warren Buffet's salary have? His salary is just a legal device. He can spend as much as he wants, although I understand that he lives relatively frugal for a mega-rich guy (or at least he used to...I don't know if he still does or not).

[The Central Scrutinizer]

Quote:

Originally posted by TeaBag420

Of course Warren Buffet, the man Bob calls the Sage of Omaha, SELLS SHARES IN BERKSHIRE HATHAWAY.

False.

[Walter Wayne]

Quote:

Originally posted by Drooper
The only way that could lead to a higher expected return, compared with buying a constant number of shares is if a lower share price was associated with a higher expected return.

If the expected return at $1 is, say 5% and the expected return at $2 is also 5%, how can different combinations of purchases at the different prices alter the expected return?

Well, I don't think DCA works well when looking at a single share type, it makes more sense when applied to a diverse portfolio. For those with only a few bucks to spare this means funds of some sort.

If I invest in company ABC, a low share price or descended share price may be an indicator that it is going out of business.

If I invest in 3 or 4 different funds, their value will track the market's (approximately). This means that right now, you are buying a lot of shares, where as at the top of bubble, you were buying few. The only assumptions necessary for DCA to work on a reasonable portfolio is 1) the markets don't completely tank over the long term 2) the markets fluctuate around the long term trend.

I think if you are poor enough that you can only go with one or two funds, DCA may not work, but if you can invest in 3 or 4 monthly, even a mininmal amount in each you will do well.

Walt

[TeaBag420]

Quote:

Originally posted by The Central Scrutinizer


False.

Okay, mmmm.... Berkshire Hathaway sells shares in Berkshire Hathaway? Or perhaps they are no longer for sale (since they have always been way out of my price range, I have never attempted to buy in).

The loins and the groins are soon dispersed.

[TeaBag420]

[quote]Originally posted by Number Six
As far as the website with the guy telling you how to get rich by market timing, you don't have to read it to know it's very likely to be BS.

Quote:

Correct. But it would help if you wanted to DISCUSS it.

Quote:

This is like standing in line at the grocery store and seeing the little booklets for say for 49 cents that have lucky lottery numbers.

This statement is quite simply insane. It's far more like those 25 cent porn movie booths. You know what I'm talking about. <---- O.K., ad hominem, but you're just being ridiculous. Continually saying you're right about something you haven't even read is, to say the least, deficient.

Quote:

If you can beat the lottery then you play the lottery, not try to sell numbers to other people. Similarly, if you can time the market you time the market, not try to make money off of telling other people how to time the market.

And when the stock market turns into a lottery, that will be a valid point.

Quote:

Keeping a radio show on the side isn't what they mean when they say diversification. Rather that means to diversity _within_ the market. If you could time the market you wouldn't need a radio show outside it. For that matter, if you coudl time the market you wouldn't even need diversification _within_ the market. All you'd need would be one aspect of the market that you could consistently time.

I know what diversification is. I explicitly said "keep your day job". I might as well have said "keep a hobby you enjoy."

Timing the market most definitely requires diversification within the market. You simply don't know what you are talking about.

Quote:

And it can't be that that guy just wants to spread happiness by having more people use his method and getting everyone rich because by definition the more people that use any method the less effective that method is.

Prove it.

Quote:

And what relevance does Warren Buffet's salary have? His salary is just a legal device. He can spend as much as he wants, although I understand that he lives relatively frugal for a mega-rich guy (or at least he used to...I don't know if he still does or not).

The relevance wasn't his salary, it was that I knew his salary, indicating that I may have done a little more research.

Game. Set. Match.

[Number Six]

ME:
If you can beat the lottery then you play the lottery, not try to sell numbers to other people. Similarly, if you can time the market you time the market, not try to make money off of telling other people how to time the market.

YOU:
And when the stock market turns into a lottery, that will be a valid point.


ME: For the purposes of this example, the stock market is the same as the lottery. In both cases, if you have a way to beat the system then you can make a lot more money by challenging the system than by selling your system to others. OTOH, if you can't beat the system then you can make a lot more money by selling your system to others than by challenging the system.

----------------------------------------------------------------------

ME:
Keeping a radio show on the side isn't what they mean when they say diversification. Rather that means to diversity _within_ the market. If you could time the market you wouldn't need a radio show outside it. For that matter, if you coudl time the market you wouldn't even need diversification _within_ the market. All you'd need would be one aspect of the market that you could consistently time.

YOU:
Timing the market most definitely requires diversification within the market. You simply don't know what you are talking about.

ME:
No, it doesn't. If you consistently can time any single aspect of the market you can get rich quickly. If you can always tell when bonds are going to go up or down then what difference does it make if you can tell what will happen in any other aspect of the market? All you have to do is play the part of the market that you can accurately time and ignore the parts of the market that you can't accurately time.

-------------------------------------------------------------------

ME:
And it can't be that that guy just wants to spread happiness by having more people use his method and getting everyone rich because by definition the more people that use any method the less effective that method is.

YOU:
Prove it.


ME: Suppose there was a great system that you had and nobody else had. The system says that under conditions A, B and C you buy and under conditions X, Y and Z you sell. And you always win. Fine.

Next suppose you tell your system to me. Now, both you and I buy under conditions A, B and C and sell under conditions X, Y and Z. Consequently, demand very slightly increases under conditions A, B and C compared to when only you used the system. Also, demand very slightly decreases under conditions X, Y and Z compared to when only you used the system.

Then suppose I tell a third person about our system. Again, demand very slightly increases under conditions A, B and C compared to when just you and I used the system. Also, demand
very slightly decreases under conditions X, Y and Z compared to when just you and I used the system.

As more people learn about the system, demand under conditions A, B and C increases and the demand under conditions X, Y and Z decreases. Since you're buying under conditions A, B and C, as more people learn the system there is greater demand under A, B and C, which means you have to pay more when you buy. Since you're selling under conditions X, Y and Z, as more people learn the system there is less demand under X, Y and C, which means that you get less when you sell. As a result, the more people that use the system, the less money each makes.

Suppose you take it all the way...more and more people use the system until eventually everyone uses it. If everyone uses it then everyone has the same conditions for buying or selling. And if everyone has the same conditions for buying and selling then nobody ever buys or sells and the market comes to a stop (in which case nobody makes any money).

[Ziggurat]

Quote:

Originally posted by Drooper

By using this "DCA" all you are doing is allowing the number of shares purchased with each tranche to vary.

Yes. Which weighs the cost average more heavily towards the cheaper shares, compared to share cost averaging.

Quote:

The only way that could lead to a higher expected return, compared with buying a constant number of shares is if a lower share price was associated with a higher expected return.

If the expected return at $1 is, say 5% and the expected return at $2 is also 5%, how can different combinations of purchases at the different prices alter the expected return?

You're kidding, right? If you bought a share in the same company at $1 once and at $2 the next day, it doesn't matter what the expected returns are, the REAL returns are obviously NOT going to be the same. The math doesn't lie, and I showed you exactly how the math works.

Quote:

You are using this DCA as an implicit contra-indicator of expected return, by saying that by buying more shares when the price goes down increases your expected return, which is wrong unless you have some information that the market is not pricing.

Nonsense. You're confusing buying more shares when the cost is low than when its high (DCA) with market timing, where you buy shares in *response* to a drop in the price. The later is market timing, and will not work. The former is not timing, but merely ensures that the averaging between prices comes out in your favor compared to share cost averaging. Do the math: the numbers don't lie. Expected return != real return.

Edited to fix format.