Dollar Cost Average During A Bear

[Patrick]

Well I am not smart enough to figure this out with some simple computer spread sheet--just pen and paper so very slow BUT so far-I have done a little data checking--and as I suspected--most of the time the market rarely has more than a few months in a row of sequential losses...from 1950 to 2000 there were a few dry spells- in the mid 60's there was a 6 in a row, and in the 70's an 8 in a row, followed shortly thereafter by a 9 in a row skid...BUT as I suspected IF - and it is a big IF--IF you have unlimited ability to increase the bet a progressive system does generally win more than it loses....particularly with a market where you do not generally lose all of your bet...but even in an all or nothing ROULETTE WHEEL- contrary to what Ha is saying it is a statistical fact that more often than not you win --you don't win a lot but wins are more common than losses....BUT when you lose you can lose HUGE!!!! You have to win more because in any 50 50 scenario the winning and losing has to be equal--and we all agree that there is at least a small chance of many losses in a row and a huge loss--so if there is a small chance of losing huge then there has to be a large chance of winning a little to make it all balance out--it is this imbalance and the need to have huge amounts for the rare bad streak (and the limits on large bets) which keep anyone from actually succeeding at betting progressively in any casino....it is not beating the odds--the odds say that in a game with 50/50 chance (not quite accurate with roulette because of the green numbers) you will win as much as you lose (MONEY WISE) but if you could bet unlimitedly and progressively - you will be up more often than you are down--but when you are down you will be down very very big...
With investing it is different because it is not a win all lose all and there are dividends, etc...BUT if you started progressively investing in the S &P 500 at the beginning of the WORST streak in the last 60 years (December 1922 to Nov 1, 1974) and started with 1 unit investment- you would have had to sunk 512X your initial investment in once in the last BAD month of a 9 in a row streak--but the next month you would be up over all by about 2.5%...but you would be up and if you never sunk another nickel in-because you shot your wad -- a year later you would be up 30% total...not including dividends in any of this calculation
ON the other hand if you invested the exact same amount in equal measure over the same time frame you would be down at the end 22% and if you never put in another nickel, because you shot your wad -you would still be down 3.8%.
Now the next calc-which I have not yet done is to see what happens when you do this in a good mostly up market--because the even investments are going to be bigger in the beginning and therefore benefit more than the more timid progressive inputs-in fact less is going to go in because you cannot know how much you will be putting in ahead of the fact--still I can't help thinking that a general increase after a bad month, whatever you can afford and a return to baseline after up months might not be a slightly advantageous inputting strategy in a market that in the long run is expected to go up. Unlike dice and roulette wheels- the market does have a memory and a trajectory.

A few questions so I can understand how your strategy is supposed to work:

What is your cash percentage at the beginning?

When will you sell stocks to replenish the cash?

What is your cash percentage after having sold the stocks so you can begin to purchase stocks again?

PS - on an unrigged roulette wheel you do not win more often than you lose. You lose more often than you win. Even in a 50/50 game of chance you win 50% of the time, therefore you lose 50% of the time. If you win more than 50% of the time, it's not a 50/50 game.

 

[GregLee]

BUT when you lose you can lose HUGE!!!!

I have some doubt about whether the HUGE losses will really happen. Maybe such losses would prevented politically, because they would be damaging to too many voters.

 

[ERD50]

Well I am not smart enough to figure this out with some simple computer spread sheet--just pen and paper so very slow BUT so far-I have done a little data checking--

Patrick asked the Q's I'd be asking, I'll just point this out, which is just another angle of what Patrick was getting at (What is your cash percentage at the beginning?):

.... - you would have had to sunk 512X your initial investment in once in the last BAD month of a 9 in a row streak--but the next month ...

OK, so if I go with your numbers (they are probably close), how the heck do I come up with 512x my initial investment? Using margin, I would still need ~ 256x. That seems to lead to only having 1/256 of my initial investment in the market, and the other 255/256ths in something stable. So I'd miss out on a lot of market rise since I'm mostly out of the market.

And if even a subset of this would goose earnings, I'll go back to my earlier statement (which I apply to almost everything of this nature) - if it works, why isn't some mutual fund using the technique and reporting consistent above market returns? I'd guess that is because it doesn't work.

Not trying to burst your bubble, not trying to e negative, I'm actually interested in anything that might beat the market. But I'm always skeptical, and that has saved me a lot of money over the years.

-ERD50

 

[urn2bfree]

I appreciate the feedback and questions and need to clarify a few things.
1) THESE ARE THOUGHTS FOR A SYSTEM FOR THE ACCUMULATION PHASEONLY-- not withdrawl--the cash comes from income- from your cash--this is strictly a BUY AND HOLD and has nothing to do with stocks vs bonds - it is cash vs invested-
2)It is not realistic to have the kind of liquidity to fund a truly PROGRESSIVE system
(ok quick final digression on this--for those who still dont get it--PROGRESSIVE systems work to allow you to have more wins of small amounts at the risk of fewer loses of large amounts- IN a true 50/50 system, if you always bet the same amount you will have equal numbers of wins and losses and equal amounts of wins and losses so it all zeros out. IF you vary the bets to be only 1 unit at start and stay with 1 unit with every win, and double the bet after every loss, you will have equal amounts won and lost but you will have more number of wins of small amounts and fewer losses of large amounts so it all evens out...-- every time you lose in a row the losses mounts quickly--but if you win just once you are back up one unit more--the odds of winning once are always 50/50 but the odds of losing many times in a row are much lower than 50/50-- you will win more OFTEN- just not the large amounts that you will risk on those rarer occasions of multiple losses- Put W as a win and L as a loss-
in a simple betting system without progression--bets of equal amounts on each spin- LWLWLWLWLW- 5 losses, 5 Wins- you end up even.
But in a progressive system the exact same sequence of LWLWLWLWLW and you are up 5 units --and any sequence that ends with W (there should be half of them) you will be up the number of W units in the sequence...Even in the very rare LLLLLLLLLW sequence of 10 spins you end upwith a gain of 1-- if that were the sequence in a standard betting system you would be down 8--
...only sequences that end in an L and have multiple L's ending in that L will lose and that happens much less than half the time...but when it does you can be down big....if the above 5 L and W is WWWWLLLLLW you end up 5 just the same-but just before that 10th spin you were down 27 units and have to risk 32 more units to get back to up 5.)
3)SO doubling the investment each month is not only not realistic but requires you to keep cash on the sidelines in case of the rare multiple downs and misses out on in an up market...
4a)markets are NOT roulette wheels, if you believe in market efficiency you will win more than you lose and the cash you invest in stocks and bonds will produce in the long run and not even out...the key is to buy more when low and less when high and go long (we are not talking about selling here at all)
4b)markets are not roulette wheels- what is going up will come down and what goes down will come back up and corrections and rebounds are inevitable
5a)I am thinking the percent changes per unit time can inform the investment- rather than always putting in the same amount--
if I normally put $1000 a month in to 60% stocks and 40% bonds maybe if I vary the amount by the monthly return of the portfolio as follows- if the month is even- I put in $1000, if the monthly return is down say 1.2% I put in 1.2% more...if the next month the return is down even more -I increase the $1000 by that same percentage -if it comes up, I decrease by that percentage increase--if there are many up months, the investment gets smaller and smaller each up month--sure you miss out on some of a bull market, but many up months seems like inevitably a correction is coming and my input would be smallest at the top...likewise after multiple down months, my inputs would be bigger and bigger towards the bottom.
5b)seems to me the percentages up and down should be measured against different baselines--that is if going down adjust by the amount down from the beginning of the slide--but on the upside adjust only according to the prior month or two...
6) By doing this on a more conservative % basis rather than progressive doubling the liquidity risks are lower...I would have to come up with less drastic cash infusions...and liquidity is built up by the surpluses when I am lowering my input in the up months.

PS- it does not have to be monthly adjustments/investments-it could be quarterly or yearly--the idea is that putting in the same in a traditional Dollar Cost Averaging manner could maybe be tweaked by conditions at each investment period...I chose months.

 

[Senin]

Wow. This thread has certainly gone far afield from my original idea. But, hey, it is a free country.

I still like the idea of keep a small reserve and in a down market, buy buy buy.

 

[Patrick]

Urn2Bfree, if your investment money is all coming from income, how do you get 1.2 or 1.5 or 2 times more money to invest after a down month? I'm thinking that you must have withheld some in a previous month that could have been invested. So if the market goes up instead of down, you have cash on the sidelines that is not invested.

If you now change your allotment (after the market has gone up some), you have purchased at a higher value than you could have if you just invested the money when you got it from your income stream. If the market now goes down, you have a paper loss from that money and now your system says to put more in at a price that you could have paid earlier with the withheld cash.

I suppose if you have the correct sequence of market gyrations that your system would indeed result in an overall gain. The challenge is designing the system to fit the gyrations, then having the luck for the gyrations to occur in the correct sequence.

 

[ERD50]

... (ok quick final digression on this--for those who still dont get it--

PS- it does not have to be monthly adjustments/investments-it could be quarterly or yearly--the idea is that putting in the same in a traditional Dollar Cost Averaging manner could maybe be tweaked by conditions at each investment period...I chose months.

Back test it. Report back. Everything else is just talk. - ERD50

 

[urn2bfree]

Well- as usual the best thing to do is keep it simple- and I confess I got lost making this way more complicated than it needs to be-what I am actually wrestling with here is what should be the percent of your total portfolio kept in cash--up to now my cash position has been based on hedging against unemployment--ie so many months of expenses... I have a steady stream of income and I just keep the back up cash to cover expenses for several months and put the rest in...but is that best? I dont think expenses should determine allocation percentages--but I just don't know what the answer is...still seems to me that more cash in when you are down and less cash when you are up is just keeping the percent allocated to cash stable.
---so first there is the AA of invested assets--I like the middle of the road assertiveness of 60% stock 40% bonds--but that is not taking into account the cash--as I accumulate cash savings how much should I keep?...not based on expenses but to be ready to move-- I don't know the ideal "shape" to be in to be able to exploit market fluctuations.
so what is ideal?
57% stocks, 38% bonds and 5% cash
54/36/10
51/34/15
48/32/20
I am sure someone has analyze this to find the optimal growth with the least volatility....

 

[GregLee]

corrections and rebounds are inevitable

Why do you think that?

 

[Patrick]

Well- as usual the best thing to do is keep it simple- and I confess I got lost making this way more complicated than it needs to be-what I am actually wrestling with here is what should be the percent of your total portfolio kept in cash--up to now my cash position has been based on hedging against unemployment--ie so many months of expenses... I have a steady stream of income and I just keep the back up cash to cover expenses for several months and put the rest in...but is that best? I dont think expenses should determine allocation percentages--but I just don't know what the answer is...still seems to me that more cash in when you are down and less cash when you are up is just keeping the percent allocated to cash stable.
---so first there is the AA of invested assets--I like the middle of the road assertiveness of 60% stock 40% bonds--but that is not taking into account the cash--as I accumulate cash savings how much should I keep?...not based on expenses but to be ready to move-- I don't know the ideal "shape" to be in to be able to exploit market fluctuations.
so what is ideal?
57% stocks, 38% bonds and 5% cash
54/36/10
51/34/15
48/32/20
I am sure someone has analyze this to find the optimal growth with the least volatility....

Since you have 40% in bonds, why not use that money to Martingale into the market while you're living on the income stream? Use the excess income to reestablish the bonds over time.

 

[urn2bfree]

The "inevitability" of corrections and rebounds is more a confidence that they will occur borne of prior experience and a necessary expectation for anyone with belief in investing and dollar cost averaging.
I found a Forbes Calculator which estimates returns on various portfolio allocations and allows you to fix the expected rates of return and for the bond and equity portions. It also has a cash allocation. With my very conservative rates of return and splitting the "at risk" part of my portfolio at 60% equity/40% bonds (But lower percents of the total when accounting for cash) I played with different cash percentages. The returns only went up a little below 10% but dropped more above 15%. My conclusion for me is the risk withlower than 10% cash is not worth the slight gain in return.. And the lost return above 12% cash is not a trade off that I like.