Drip feeding is a waste of time

TheTracker

Even drip feeds are lumps. And any lump is but one big drip. We all drip feed over a lifetime. We all make lump sum deposits.

There is a key test: Can I emotionally and intelligently cope with a drop of 40% of an amount of money I plan to contribute? If no, reduce the contribution amount until comfortable, or train your emotions to allow such a drop.

jimjames

cloud_dog wrote: »

Some people are not and may never be comfortable making lump sum investments. Investing for the long term is positive and should be encouraged, and therefore using a regular investing method is nearly as good as the big bang approach.

I'd agree. Especially as regular investing ties up quite nicely with receiving regular income too.

bowlhead99

cloud_dog wrote:

If you had said....you invest £10k for 10 years a) by single investment b) by £500 per month for 20 months; that is a more realistic scenario. And, one where the variation is likely to be relatively small (depending on market fluctuations over the period)

As the variation is likely to be relatively small between taking a random 'average' price in a tight 20-month window ahead of a 9 year investment journey as opposed to a price from a one-month window at the beginning of a ten year investing journey, there doesn't seem to have been much point dragging the investment out over almost two years. The investment might crash in month 2 or it might grow for 21 months and crash in month 22. The former would give a much better result for 'averaging' and the latter a much better result for 'lump summing'.

But I acknowledge that some people may find it emotionally difficult to commit to the logic, especially if they don't really want the risk of being in a risky fund and prefer to cling to cash for as long as possible because of an innate fear of investment risk.

Your continual reference to 'risky' is this generic or specific?

Originally Posted by bowlhead99
But investments are not binary on/off, 1/0, high risk/no risk. They are a range, a graduated scale. So, if you don't like the idea of being able to lose money on a risky investment, don't decide to blindly go ahead and buy the risky investment anyway by dripping into it, keeping most of your cash out of it for as long as possible because you really don't like the risk. Simply, buy a different investment!

You're missing the point. It's not about risk profiles of investments, it's about investing.

It is absolutely about risk profiles of investments. The only reason to take any mathematically poor decision to not invest ASAP and instead invest slowly-slowly is because you don't like the risk and would rather have a portfolio which has a risk/return profile of halfway between the risk of the investment and the risk of cash, for the next 20 months or whatever.

This can be better solved, not by refusing to buy the investment in whole and simply buying it in part by clinging on to mostly cash as an insurance policy against buying the investment... but by finding a more suitable investment.

cloud_dog wrote: »

Rather than these novels, I should have simply summed up my point....

Some people are not and may never be comfortable making lump sum investments. Investing for the long term is positive and should be encouraged, and therefore using a regular investing method is nearly as good as the big bang approach.

Yes, any long term investment is better than not at all, so I do agree it is better to be cajoled into doing it slowly than being shouted at that you have to do it all at once. If it gets them invested when otherwise they wouldn't have been, fair play. That is certainly what the investment platforms play on in their marketing as it is the only way to get some people over the line.

The original premise was:
"If you can, lump sum investments beat drip feeding in the long term." which is a position borne out by research as well as common sense.

Your response was: "it's horses for courses" implying the premise was not really true and perhaps lump sum investments did not usually beat drip feeding in the long term.

But based on your explanation, it sounds like you don't dispute that "If you can, lump sum investments beat drip feeding in the long term." is entirely true.

Because all you are saying is that "If you cannot (because of your psychologic tendencies which you are unable to overcome), lump sum investments do not beat drip feeding in the long term because the amount invested would be zero under lump sum and some positive number under drip feeding."

Glad that's all cleared up. Pint?
:beer:

cloud_dog

Sort of agree

You missed the bit off from the OPs post about saving up until you have a pot of money to invest. go back to the psychology of making an investment decision

Plus

Assuming the market drifts upwards in the long run, but with noise, my assumption would be that a lump sum investment is likely to have a higher mean, but also a higher standard deviation. In other words, on average you're better off, but with a broader spread of possible outcomes.

A regular investment might have a lower standard deviation, because you're sampling at more points which means you're more likely to achieve something near average, but you spend less time in the market so a lower mean. That means on average you'll make less than the lump sum case.

In other words, lump sum is higher risk/higher reward than a regular investment.
As you extend the timelines the noise becomes less significant so the risk reduces but the return does not.

Meanwhile, over the period of regular investment your money is in the market for half the time so you make less return for that reason. If you do regular investment for a period then hold, as time goes on that looks more similar to the lump sum investment.

That's my intuition: I haven't run any numbers to test it.

redux

cloud_dog wrote: »

Sort of agree

You missed the bit off from the OPs post about saving up until you have a pot of money to invest. go back to the psychology of making an investment decision

And there seems to be a clash between these two remarks.

bigfreddiel wrote: »

Drip feeding can be costly in a couple of ways, the cost of purchasing an investment, and the loss of dividends because the entire sum is not invested.
_

Of course you may not have the funds to make a decent lump sum investment, but you could use a 5% savings account to build up your lump sum.

If someone saves up cash first in order to make a lump sum investment, they are more likely to miss a dividend than the person making monthly investments straight into the fund.

bowlhead99

Plus wrote: »

In other words, lump sum is higher risk/higher reward than a regular investment.

Effectively, that has to be the case.

In regular investment you are getting yourself a blended average start price by mixing together the risks and returns of your chosen investment with the risks and returns of cash, for the entire period while you're dripping.

As the risks and rewards of cash are lower than investment, it has to be true that drip feeding is lower risk and lower reward than investment; your average risks and the average return of which you're capable, is reduced.

Which can also be rearranged to say lump sum is higher risk / higher reward, and this must be the case, because rewards are only available to those taking risks, and if your money is still in cash then you are not taking the risk. So you won't get the rewards - Disregarding 5% regular saver accounts

As you extend the timelines the noise becomes less significant so the risk reduces but the return does not.

Meanwhile, over the period of regular investment your money is in the market for half the time so you make less return for that reason. If you do regular investment for a period then hold, as time goes on that looks more similar to the lump sum investment.

That's right. In the period when you're fully invested you get the full returns, and when you're not you don't. So if you drip fed for five years and left for 45, versus just investing on day one and leaving for 50, the shape of the annual performance graphs would track each other exactly in the last 90% of the graph, and you could say they look more similar.

Your comment about "As you extend the timelines the noise becomes less significant so the risk reduces but the return does not" might mislead some people into thinking that the graphs will basically converge and they should probably just take the slow cautious way to invest on a drip feed and still get basically the same result. Just to clarify for readers, this is not really the case.

Of course, if as a result of the statistically higher returns in the first five years, you went into year 6 with a £11k pot from lumpsumming instead of a £10k one from dripfeeding, and then got your 7% annualised for 45 years...

... then by the time you get to the end of the 50 years your 2000% profit would leave you with 'only' £210k in the dripfeed portfolio and £231k in the lumpsum. The difference in the pot size that you lock in by picking one method at the beginning, persists - so that if lumpsumming got you an extra 10% then that differential of 10% in pot sizes will continue forever.

That extra £1k of gain from taking the returns of the investment over the first five years rather than the returns of an investment-and-cash blend over the first five years, will snowball into an extra £21k of gain which might be very much appreciated. The problem is that you can't know you are going to get the gain, because you can't know that investment will outperform half-investment- half-cash in that specific five year period, even though, statistically, investments do outperform cash over five year periods. You might instead make less money doing it that way. It is only probable and there can be significant variance / standard deviation to the results, as you say.

That's my intuition: I haven't run any numbers to test it.

It's pretty much common sense and maths really, so although Bigfreddie mentions credible research proving the data effects over long enough time periods, which will use real-world data in all shapes and sizes rather than neat theoretical examples... it's probably not necessary to actually run the data.