Keeping one's powder dry

[Adamski]

Unless you're an investing genius, can't time the market. The best time to buy is when people are on here having a melt down, selling their entire portfolios, like a few did in March 2020.

But then when it looks terrible take a brave person to convert savings from cash to shares when everything is in the red. And more recently in 2022 Tesla dropped from 300$ to 100$, and beginning of year I thought still going to go down, but stock price has reversed and back up to 263$. Difficult to find the logic there, .. the madness of crowds.

So I think best way is deciding on a long term strategy 10 years, and sticking with it. For me it's a 60:40 stocks/bonds. Others it's 100% in the market. Depends what you want and your risk tolerance.

[Gerry557]

Personally I do try and keep dry powder available. It wasn't always the case. Initially I invested a set sum monthly. As time and the size of my PF increased and I got a feel for shares I started to invest in tranches. I would try and time the buys with dips.

Sometimes your PF does well and the funds available grows as its not reinvested. Then some might get invested just to keep things growing.

It's also handy to have some cash for emergencies anyway. I also keep access to extra funds but marked more for black swan events when things can get silly.

Obviously after dips or swans the cash needs to be restocked for the next time.

There is a counter argument. I read an article about a bad investor who invested at the height of each bubble only for the market to drop significantly just after.

The result was he did fine over the long period he was invested. I think it was on the Fidelity website.

https://endowus.com/insights/worst-inve ... d-strategy

[vand]

Even if you can execute it perfectly, the value added by being able to time the bottoms is only marginal over the long run:

https://ofdollarsanddata.com/even-god-c ... averaging/

[spasmodicus]

Even if you can execute it perfectly, the value added by being able to time the bottoms is only marginal over the long run:

https://ofdollarsanddata.com/even-god-c ... averaging/

Yeah, that's as may be, but your reference says

The only other rule in this game is that you cannot move in and out of stocks. Once you make a purchase, you hold those stocks until the end of the time period.

So the "buy on the dip" expert isn't allowed to sell at the peak? That's a crazy game strategy. Personally, I have a bigger problem than whether to keep my powder dry, or keep it a bit damp by putting it in gilts. I have always had a problem with knowing when to detonate my powder, i.e. to de-invest or sell, which applies just as much to unsuccessful investments as to the ones that shine (struggling here to find a "powder" metaphor appropriate for selling).
There are strategies out there that persistently beat "pound cost averageing", e.g. see Robert Lichello's book
"How to make $1,000,000 in the stock market". 4th edition, Penguin.
Explosively,
S

[simoan]

As I understand it, 'keeping one's powder dry' is strategically holding cash until prices become more favourable. I've been holding new cash in my ISA waiting for a correction in the price of a global tracker (iShares World ETF SWDA). In March it was around 6100p, increasing in April to around 6200p, and I've been waiting for a correction back to 6100p since then. It's now over 6700p. Am I keeping my powder dry, or am I just trying (and failing) to time the market?

You’re not doing either. What you’re doing is called “price anchoring”. It’s just about the worst mindset to get into, but it’s easy to do and very common amongst investors. Your approach makes no sense given that the investment you are considering is an index tracker rather than an individual company where you are using some valuation metric to make a buy decision. If you’re not confident making a single lump sum investment, then just make multiple smaller monthly investments over a period of time.

All the best, Si

[vand]

Even if you can execute it perfectly, the value added by being able to time the bottoms is only marginal over the long run:

https://ofdollarsanddata.com/even-god-c ... averaging/

Yeah, that's as may be, but your reference says

The only other rule in this game is that you cannot move in and out of stocks. Once you make a purchase, you hold those stocks until the end of the time period.

So the "buy on the dip" expert isn't allowed to sell at the peak? That's a crazy game strategy. Personally, I have a bigger problem than whether to keep my powder dry, or keep it a bit damp by putting it in gilts. I have always had a problem with knowing when to detonate my powder, i.e. to de-invest or sell, which applies just as much to unsuccessful investments as to the ones that shine (struggling here to find a "powder" metaphor appropriate for selling).
There are strategies out there that persistently beat "pound cost averageing", e.g. see Robert Lichello's book
"How to make $1,000,000 in the stock market". 4th edition, Penguin.
Explosively,
S

Yes. It's called lump sum investing.

But if you are talking about value cost averaging or some derivative thereof... sure, it can beat PCA but is not practical to do once your portfolio grows to a point where you monthly contributions can't compensate for the amount extra that you are support to buy on the dips.

Anyway, my viewpoint is constant - if you are going to try timing the market they you better be damn sure that valuations are in the 3 standard deviation or higher - ie, once a decade or so. And given than we had already been through 1 year of continual falls by the start of 2023 that is a hard viewpoint to defend.

[spasmodicus]

Yeah, that's as may be, but your reference says


So the "buy on the dip" expert isn't allowed to sell at the peak? That's a crazy game strategy. Personally, I have a bigger problem than whether to keep my powder dry, or keep it a bit damp by putting it in gilts. I have always had a problem with knowing when to detonate my powder, i.e. to de-invest or sell, which applies just as much to unsuccessful investments as to the ones that shine (struggling here to find a "powder" metaphor appropriate for selling).
There are strategies out there that persistently beat "pound cost averageing", e.g. see Robert Lichello's book
"How to make $1,000,000 in the stock market". 4th edition, Penguin.
Explosively,
S

Yes. It's called lump sum investing.

But if you are talking about value cost averaging or some derivative thereof... sure, it can beat PCA but is not practical to do once your portfolio grows to a point where you monthly contributions can't compensate for the amount extra that you are support to buy on the dips.

The whole point about Lichello's book is that some cash is held back for for exactly the reason that you state. In holding back cash there is obviously a risk reward tradeoff. Nobody should end up without cash at the top of the market, but how to know when is the top of the market? However, a well diversified portfolio will contain some cash-like(bonds) or negatively-correlated elements(gold) in it, which might mitigate the need to hold cash.

There is some discussion here about Licello's method here
https://toughnickel.com/personal-finance/A-Multi-ETF-Investment-Strategy-Using-Robert-Lichellos-Automatic-Investment-Management-AIM-System
I did some backtesting on this and agree broadly with the conclusions in the above. But it's maybe more appropriate to discuss that over on the "Technical Analysis" (aka "Pariahs") board.
S

[1nvest]

The whole point about Lichello's book is that some cash is held back for for exactly the reason that you state. In holding back cash there is obviously a risk reward tradeoff. Nobody should end up without cash at the top of the market, but how to know when is the top of the market? However, a well diversified portfolio will contain some cash-like(bonds) or negatively-correlated elements(gold) in it, which might mitigate the need to hold cash.

There is some discussion here about Licello's method here
https://toughnickel.com/personal-finance/A-Multi-ETF-Investment-Strategy-Using-Robert-Lichellos-Automatic-Investment-Management-AIM-System
I did some backtesting on this and agree broadly with the conclusions in the above. But it's maybe more appropriate to discuss that over on the "Technical Analysis" (aka "Pariahs") board.

At retirement, start with thirds US$ invested in US stock, gold, cash - split 50/50 hard cash and short term Gilts. Draw a DIY 3% dividend yield (3% SWR) and you've eliminated early years bad sequence of returns risk. Held as-is for 30 years and that's inclined to end with at least half your inflation adjusted portfolio value still remaining. Supplement that with some timed additional stock purchasing and the overall mid/longer term rewards can be better than that of having started/held 100% stock.

Much of stock 'average' gains are a consequence of skewedness. In around 75% of start years, 30 year windows, all-stock achieved similar outcome to that of less aggressive. Of the 25% other cases when stocks did very well (and skewed the broad average), the start dates clustered into four sets since 1871; The 1870's, around the Wall Street Crash, around WW2, and around the 1980's - high inflation low stock/bond prices lows. Most investors, >90% who lump into stock-heavy are inclined to achieve little different to more conservative asset allocations.

Buy and hold is no different to the costless lumping into stocks each/every day. If at retirement you moved into the above (very) conservative asset allocation, and started a Robert Lichello AIM applied to the stock index, price only, ignore cash interest, using that purely as a indicator of when to buy (add) to stocks, run on paper only, but where at each buy signal you upped your target asset allocation weightings by 11%, then from a initial 33% stock after 6 AIM indicated buy trades you would have moved through 33%. 44%. 55% .... 100% stock. Where typically the average cost of stock was bought at a decent discount.

Your left tail risk having been reduced, with a longer term right tail case outcome prospect uplifted.

Holding dry-powder is good only if you deploy it at appropriate times. Most unlikely you'll perfectly time the deployment, however you can still do (very) well without the precision. From such a low initial risk exposure stance, transitioning to a all-stock where the stock was averaged into at a decent discount (AIM will typically flag buys after around a 25% or greater pull-back) and there's a good prospect that you will achieve actual total returns that mid/longer term were higher than another who opted to lump all-in on stock at a arbitrary start date.

For a major stock index AIM will infrequently flag buy signals, you have to be patient, years not months or weeks. But it aids in otherwise having to make your own timing decisions, which are fraught with human emotions risk. With AIM you can pre-calculate its next trade price points, such that you can even just leave good till cancelled limit orders in place in the market, and perhaps one day/week be surprised to see that was filled at some point, in some cases where otherwise manually monitored and that opportunity would have been missed, such as having been filled via a brief market price dip that triggered your GTC.

BTW Peter Ponzo, now sadly deceased, has his summary of AIM preserved here : https://www.gummystuff.org/AIM.htm

[1nvest]

As a guide/feel, referencing AIM of BRK shares and the last AIM indicated buy occurred 11 months ago, and has since risen at a 20% annualised rate. For FCIT the last AIM buy signal occurred at the end of September 2020, since which FCIT has annualised at a 14.4% (total return) rate. More generally/broadly and 10% annualised real (after inflation) rates of return are pretty 'normal' for medium term measures, that tails off a little the longer out you go. i.e. generally Lichello's AIM does a reasonably decent job of flagging appropriate times to add to stock.

Conventional AIM also looks to 'reduce-high', replenish dry-powder, however if you just bottom draw its purchased stock the mid/longer term rewards tend to be good-enough.

[JohnW]

Holding ‘dry powder’ gives you an ‘x/y’ (stocks/cash) portfolio which should be less volatile (less risky) than an all stocks ‘x/0cash’ portfolio. Less risky, one might expect it to have lower returns, which is not a bad deal overall.

[1nvest]

Holding ‘dry powder’ gives you an ‘x/y’ (stocks/cash) portfolio which should be less volatile (less risky) than an all stocks ‘x/0cash’ portfolio. Less risky, one might expect it to have lower returns, which is not a bad deal overall.

If you start with 33/67 stock/cash and transition to 100/0 within the first 10 years, and then hold 100/0 as is for a further 20 years, you average 67/33 stock cash for 10 years, 100/0 for 20 years, averaged close to 90/10 stock/cash (88.9/11.1) over the 30 years. The difference between 90/10 and 100/0 broadly tend to be trivial.

Much of stock 'above average' total returns compared to other assets is down to less frequent great gains, such as starting after recent large declines yielding great outcomes. Unless you happen to lump 100% into stocks at such times (low probability/infrequent) your 30 year outcome rewards are inclined to be mediocre, not much different to less aggressive stock/bond asset allocations. If after having started with 33/67 stock/cash you average in the remainder 67% as pull-backs occur, such as indicated be Robert Lichello's AIM, that tends to start buy signals after 25% pull-backs, then there's a reasonable prospect that you'll end up having achieved a better overall outcome than having lumped all-in at a single point in time.

In starting with a pretty defensive asset allocation such as 33/67, if a bad earlier years sequence of returns occur, then you're far better placed than having lumped all-in.

Basically that's what Buffett suggests and somewhat does. He proposes 90/10, and having averaged in over a period such as 10 years.

You can reduce risk (such as bad earlier years sequence of returns risk), and be more inclined to achieve closer to average stock rewards (that are skewed by outlier exceptional cases), rather than a lump sum that is more inclined to lag the (outliers skewed) 'average'.

[JohnW]

If you start with 33/67 stock/cash and transition to 100/0 within the first 10 years, and then hold 100/0 as is for a further 20 years,’

If you do that you’re not holding dry powder for two thirds of the time. Different case surely?

‘Unless you happen to lump 100% into stocks at such times (low probability/infrequent) your 30 year outcome rewards are inclined to be mediocre, not much different to less aggressive stock/bond asset allocations.’

If you lump into 100% stocks your 30 year returns are inclined to be the market returns for stocks surely? The returns could be more or less than that, but the returns would be inclined to market returns I would think. And that would likely be very much different from a less aggressive stock/bond allocation like 30/70.

‘ starting with a pretty defensive asset allocation such as 33/67, if a bad earlier years sequence of returns occur, then you're far better placed than having lumped all-in.’

And conversely, if you lump sum in to 100% and you get a good sequence of returns, you’re laughing.

You can reduce risk (such as bad earlier years sequence of returns risk), (by not lumpsuming)

Yes

and be more inclined to achieve closer to average stock rewards (that are skewed by outlier exceptional cases), rather than a lump sum that is more inclined to lag the (outliers skewed) ‘average’.’

How can you, on average, get anything but stock returns with 100% stocks? Or get closer to average stock returns holding less stock and more cash, other than by imaging particular circumstances will befall you rather than viewing market movements in toto?

[vand]

My view is that if you have to make a case for yourself to cost average in over lump summing, then you asset allocation is almost certainly wrong, and you're holding more risk than you are ideally comfortable with.

Lump summing into a more balanced portfolio will beat cost-averaging with a full stock portfolio for most people, over most likely scenarios, over all but very long horizons.

Nick Maguilli has written several times about this

https://ofdollarsanddata.com/dollar-cos ... -lump-sum/
https://ofdollarsanddata.com/the-cost-of-waiting/

[1nvest]

If you start with 33/67 stock/cash and transition to 100/0 within the first 10 years, and then hold 100/0 as is for a further 20 years,’

If you do that you’re not holding dry powder for two thirds of the time. Different case surely?

‘Unless you happen to lump 100% into stocks at such times (low probability/infrequent) your 30 year outcome rewards are inclined to be mediocre, not much different to less aggressive stock/bond asset allocations.’

If you lump into 100% stocks your 30 year returns are inclined to be the market returns for stocks surely? The returns could be more or less than that, but the returns would be inclined to market returns I would think. And that would likely be very much different from a less aggressive stock/bond allocation like 30/70.

‘ starting with a pretty defensive asset allocation such as 33/67, if a bad earlier years sequence of returns occur, then you're far better placed than having lumped all-in.’

And conversely, if you lump sum in to 100% and you get a good sequence of returns, you’re laughing.

You can reduce risk (such as bad earlier years sequence of returns risk), (by not lumpsuming)

Yes

and be more inclined to achieve closer to average stock rewards (that are skewed by outlier exceptional cases), rather than a lump sum that is more inclined to lag the (outliers skewed) ‘average’.’

How can you, on average, get anything but stock returns with 100% stocks? Or get closer to average stock returns holding less stock and more cash, other than by imaging particular circumstances will befall you rather than viewing market movements in toto?

For most 30 year periods all stock outcomes are little different to less aggressive. Stock higher average rewards have elements of skewedness/kurtois, four sets of clustered start date ranges since 1870 where, following large declines such as the Wall Street Crash, stocks gains were exceptional. < 90% of investors wont however we starting at such lows, those that had the opportunity were more inclined not to (fear of further declines).

Starting with 33%, transitioning to 100% during the first decade = 66% for the first decade, 100% for the next two decades, overall 30 year average = 88.7%. If you buy the dips during the first 10 years, say 25% pull-backs, then having had 66% of your holding bought at a 25% discount yields higher longer term rewards. Not lumping in also reduces bad earlier years sequence of returns risk.

Dry powder is of course kept dry for when its needed. If you never intend to deploy it ... then it's just a burden.

90/10 where stock is averaged into over a period of no less than 10 years is what Buffett advocates and is a model he practices. Much of longer term rewards are a function of the price paid for stock. Lumping in has higher volatility, lower worst cases, higher best cases, but where most are not inclined to start with all-stock following large declines.

[JohnW]

We’re fortunate that someone bothered to model this using historical data. And yes, aggressive and slightly less aggressive portfolios have similar outcomes, but usually favour lumsumming, although the 1930’s period seemed to favour ‘wait for the drop’, for a while.

‘. If you buy the dips during the first 10 years, say 25% pull-backs,’

There have been decades and longer without 25% pull-backs which might contribute to the strategy underperforming lumpsumming in the past.
Read all about it: https://www.bogleheads.org/forum/viewto ... 3#p7449774

[1nvest]

We’re fortunate that someone bothered to model this using historical data. And yes, aggressive and slightly less aggressive portfolios have similar outcomes, but usually favour lumsumming, although the 1930’s period seemed to favour ‘wait for the drop’, for a while.

‘. If you buy the dips during the first 10 years, say 25% pull-backs,’

There have been decades and longer without 25% pull-backs which might contribute to the strategy underperforming lumpsumming in the past.
Read all about it: https://www.bogleheads.org/forum/viewto ... 3#p7449774

That (linked) model isn't the same, but the observations are still similar. Averaging 90/10 whether constantly weighted or averaged into ... are inclined to yield the same outcome. The sizeable $13M and $11M type figures presented between 90/10 and 100/0 equate to being over very long periods and amount to a fractional/small annualised percentage difference. As is noted the 90/10 has the lower risk/variance ... which is the key factor here, as we're shifting and concentrating that lower variance into earlier years.

Overall averages are irrelevant to individual samples/investors, the average figure is no comfort for those in the 10% worst of cases. Simulate 100/0 ten year variance compared to 50/50 ten years and in the worst tenth percentile cases the averages were of the order $862K vs $974K (real), the worst cases were of the order 13% more/better. The likes of 4% SWR in effect reflect the worst historic case, peak to trough. When the worst case has relatively more in earlier years then so that SWR scales higher. Whilst the longer term average differences between having held 90/10 constantly or having averaged in starting at 33/67 and increased to 100/0 at ten years in, that also averages 90/10 over x years are ... have near-as the exact same reward expectancy, that in turn are only relatively little less than average case constant 100/0.

Bogleheads = Vanguard = want you to buy their products, asap. Differences presented in a manner that make actual small % differences into sizeable $$$ differences, was at least coupled in that model/example by clarification of the risk factors.

If you lump all-in today, which is no different to what buy and holders are doing every day, then to a fixed future data your rewards are fixed to the start date valuations. Today could be a relative high, or low, locking into that is ... locked in. Averaging in and that smooths the variance, good for those who otherwise might have locked in at a high, not so good for those that locked in at a low. That's not a 50/50 thing either, as stocks generally are of the order two-thirds of time high/rising, one-third low/decline, the average individual sample/individual is more inclined to have lumped in at a relative high. That is compounded by that many investors are more inclined to buy after good gains have already occurred, less likely to lump in after/during stock declines.

[vand]

with the current mini selloff I would say that if you have cash willing to be deployed then worth beginning to deploy at least some of it.

Can you do it? Or are you still playing the market timing game?? Investment strategies and outcomes are defined by what we actually do, not what we say we do..

[vand]

The whole point about Lichello's book is that some cash is held back for for exactly the reason that you state. In holding back cash there is obviously a risk reward tradeoff. Nobody should end up without cash at the top of the market, but how to know when is the top of the market? However, a well diversified portfolio will contain some cash-like(bonds) or negatively-correlated elements(gold) in it, which might mitigate the need to hold cash.

There is some discussion here about Licello's method here
https://toughnickel.com/personal-finance/A-Multi-ETF-Investment-Strategy-Using-Robert-Lichellos-Automatic-Investment-Management-AIM-System
I did some backtesting on this and agree broadly with the conclusions in the above. But it's maybe more appropriate to discuss that over on the "Technical Analysis" (aka "Pariahs") board.
S

i can't take an article seriously that only considers a 1998-2014 timeframe.

i am not saying such methods aren't worth investigating, but nearly all of them tend to fall apart going forward, case in point all these strategies that try to introduce a filter for only buying markets that are above the 200dma look great in backtesting but sucked when put into practice over the last few years.

[Itsallaguess]

With the current mini selloff I would say that if you have cash willing to be deployed then worth beginning to deploy at least some of it.

Can you do it? Or are you still playing the market timing game?? Investment strategies and outcomes are defined by what we actually do, not what we say we do..

The problem with 'keeping some powder dry' is that there's a risk of perhaps turning things into a 'black or white' situation, which then makes things difficult in determining any 'trigger points' for action based on the markets themselves.

One way around this that I've found works well for me is to set a limit on powder-capital, so that it's far enough away initially, once a 'hang-back' decision is initially made, that reaching it will definitely feel like an 'accomplished goal', but close enough that gives me a chance to then go ahead and re-participate in what might still then be a 'subdued market' with any further, additionally-available capital over and above that previously set powder-limit.

That approach has worked for me over a number of years now, and has allowed me to stay 'fully invested' with already-invested holdings, whilst still allowing me to feel like I'm sometimes actively taking a step back for a while in terms of further-invested capital, and allowing what would normally be regular top-ups to then accrue into the 'powder' pot.

Once a pre-determined level of powder-pot has been reached, I'll then allow any further accrued capital or dividends to continue to be invested in the market, maintaining that higher-level powder-pot in case I've got things totally wrong and the market continues to fall.

Cheers,

Itsallaguess

[Whatsitworth]

“Always keep one’s powder dry”

Sir Arthur Wellesley, the duke of wellington


Or maybe it was Pablo Escobar