I just spent 2.5 hours to simulate a few scenarios about “buying the dip”, including this proposed strategy and will summarize the methods/results below. As we all know and read many times, it is overall less successful to hold money back for market timing. So if we follow this statement, there should never be the option to “give some extra” into the dip as everything is already invested (aka time in the market > timing the market).
However, reality and the human psychology works often a bit different. If something happens that we suddenly feel more motivated to go a bit “more” into the risk and invest a bit more to “buy the dip” and overperform vs. the market. From this (from a rational standpoint wrong) perspective, we can now compare some scenarios.
Scenario 1: Baseline
invest 1000$/month into VT (always after salary payment; on 28. of the month)
no additional investments/adjustments due to market performance
Scenario 2: Immediate Buy-the-Dip
follow the baseline scenario, but if there is a >=2% drop of VT in one day, then immediately invest another 1000$ on the next day
maximum 1 additional payment per month (so either we invest 1000 or sometimes 2000$); any following drops of >=2% are ignored because there is no more money available to invest
Scenario 3: Short-Delayed Buy-the-Dip
follow the baseline scenario, but if there is a >=2% drop of VT within 3 days, then invest another 1000$ on the third day
could help to avoid buying while market still goes down
Scenario 4 (inspired by @TeaGhost): Cool-Off Buy-the-Dip
follow the baseline scenario, but if there is a >=2% drop of VT on one day, it waits until at least 3 continuous days there is no more than a 0.2% downturn per day (measure of reduced volatility); if this is the case, buy in with 1000 $
if this becomes the case but the price is already higher than before the drop (e.g. one day after a -2% you get a +2.2% rebound), it ignores the event and continues with the baseline investing
max 1 additional event/month (max 2000$ investing/month)
Scenario 5: Baseline 50-50 Dip
invest 50% of the total investment amount (e.g. 500$ of 1000$) on any day during the month if there is a 2% drop in VT price
invest the remaining 50% at the end of the month as in the baseline scenario
if there is no drop during the month, invest the full 1000$ at the end of the month
every month the same amount is invested; only difference is the timing of 50% of the amount within the month
Scenario 6: Hardcore Buy-the-Dip
similar to scenario 5, but here we invest the full amount (100%, so 1000$ out of 1000$) as soon as we see a 2% dip
if there is a dip during the month, no additional payment is made at the end of the month
if there is no dip, we invest the 1000$ at the end of the month (same as in baseline scenario)
Methods
simulated over a timespan of 10 years, and starting at each month between 2008 to 2014 (and then 10 years from the start) → total nr of simulated 10-year timeframes = 120
calculate return in % for each of the 120 timeframes and for each scenario
get mean % of all timeframes and standard deviation
return = return of investment → (sum of all investments)/(final value of portfolio)*100 - 100
Results
All results here are only about % return. Absolute return varies due to different amount invested.
Note, for baseline, the return of investment in % would be the same if I invest 1000$ or e.g. 2000$/month (corresponding to 240 buy-events, but 2 on the same day; disregarding fees).
The strategies are all kind of similar (given that we stick to the baseline investing for each of them)
Small advantage in return for buy-the-dip strategies (but remember, that technically only works when the money “appears” during a crash and could not have been invested before)
Immediate Buy-the-Dip was not worse than other strategies such as cool-off or delayed buy-the-dip; but cool-off only needed 11 additional investments (vs 40 for immediate and 55 for delayed) to get the 1.5% edge over baseline (62.2 vs. 60.7)
Increased return is only visible when adding additional money to the table, while trying to time within a month did not help at all (see scenario 5 and 6)
→ likely because changing the exact day within a month is too small to make a real difference; but profiting of bad months can slightly help
Lots of assumptions went into that calculations. Given that the outcome is similar for all of them, I wouldnt invest more time digging around. But if someone wants to continue, let me know and I’ll share my Jupyter Notebook
I’ll say, though, the outcome totally depends on how you define the non-Baseline (aka non-DCA) scenarios.
I just recently came across a post that compares DCA to “Buying the Dip” where “the dip” means buying at the exact market low between any two all time highs.*
Turns out DCA beats most of the time.
* Admittedly, I feel that’s a little contrived as well, but so are the other buying the dip scenarios. At the end of the day, it’s investor psychology what investors perceive as “the dip”, and it probably not only varies from investor to investor but even within one investor over time.**
Yeah, that’s a problem, I think. Considering that markets went mostly up, you have more money compounding. Furthermore, it is a rather unrealistic scenario of money appearing when you need them . More seriously, I am not sure how you have calculated the return and if it is correct. I don’t even know what is the right way to calculate it in this scenario.
What about the following scenarios:
you always invest 1000$ per month
on the last trading day of a month (or 28th) in the base scenario or if it was not invested by a trigger
before the end of the month according to your triggers.
In this case, the total number of buys and the total money invested would be same.
Sorry, but I am not in a position to take over.
And the inflation adjustment would be another story, I am afraid.
Thanks, this is very interesting and something I always wondered.
In the buy the dip cases, you list 160 buy events versus 120 in the base case. That means 160k was invested instead of 120k which is substantially more.
I think it would then make sense to compare with a scenario where the same amount was invested every month over 10 years. In other words what happens when there is no buy the dip, but we are now investing 1333 per month (which would be close to 160k in 10 years)?
My initial reaction was similar: you just have more money overall with time in the market, hence the outperformance, but I wasn’t sure whether the potentially $2k deployed in the buying-the-dip scenarios were saved from months where no dip occurred.
If not, the simulation should probably be adjusted to deploying equal amounts of cash in each scenario?*
* As done in my referenced comparison between DCA and buying at the “perfect” dip (in the latter, cash is saved up until the “perfect” dip appears).
In my opinion, to compare apples to apples.
In each scenario , same amount of money needs to be invested
Now since you are trying to simulate market timing, the baseline needs to have uniform investment every month . But the other scenarios should have delayed Investment because one is trying to buy the dip.
So baseline -: 1000 USD invested on 28th of every month
Other scenarios -: Investor didn’t invest on 28th and kept waiting and then at some point during the month, an opportunity arrives and investor invests. If buy the dip opportunity doesn’t arise, then investor will invest on next 28th. So the longest wait would be 30 days.
Thanks for the suggestions, I’ll look into making some more calculations
Here a few answers to start with:
For this example, I calculated return in the sense of the total return of investment:
(amount of money at end of simulation timeframe)/(amount of money invested).
E.g. 120k invested, 130k in the end → 130/120 = 8.3% return
I mainly care about how much money there is at the end, and less about how much each individual buy-in made etc. Therefore I feel like this measure seems the most intuitive to me.
Both good ideas, might give them another shot.
Indeed. But not only inflation, also Fx rates. USD-CHF was much higher 10 years ago vs. today.
I actually did include this in my simulation, but since based on the pure return % you get the same outcome, I havent included it in the post. Imagine you invest every month 1000$. After 10 years you invested 120k and you got out 180k → your money increased by 50%
If you now invested 2000$/month you would have invested 240k and got 360k → still increased by 50%.
→ added a brief note about this to the post above as well
Good posts on buying the dips. The key assumption missing, in my opinion, from these analyses is whether an investor who diligently plugs money into the market at set intervals (ie not DCAing, if we want to get pedantic) manages to miracle some money up to buy dips OR is keeping “dry powder” in order to buy dips. I guess the variables are too many to do this properly.
Keeping “dry powder” we know well that is a losing game, in Peter Lynch’s words “more money has been lost waiting for corrections than in corrections”.
By “miracle some money up” I mean for example raiding their emergency cash reserves and/or dedicating to eat baked beans for 3 months, or eating the family cat instead of chicken, or other wholesome practices like that. In my own greengrocer-level mental accounting I have kept track of the “chunks” that were used to buy dips - last example is 27 October 2023 - and indeed they have done measurably better. The rub is that my own liquidity is not substantial enough for gains to make any meaningful difference, meaning that if buying a dip would net me a couple of hundred franks more than not doing it then the only reason to do it is psychological.
In other news, I can’t shake the thought for months now that Buffett is preparing for the bull run to end in a bang where he (or his successors) can swoop in and eat up businesses like a blue whale eats krill (edit: please don’t get pedantic, I know BRK isn’t interested in krill-level business, you get the analogy!), either one last hurrah, or setting up his successors for…success. I never intend to stop adding to my BRK.B position.
Another easier way is to tweak asset allocations. If you have, say, 70/30 stocks/bond, then this automatically buys the dips if you re-balance on a dip. Or you can buy the dip by shifting the allocation 75/25 etc.
Hussman’s analysis hypothetical swapping between stocks and T-bills based on market internals:
As far as I understand, T-Bills had negative returns in 2021 - Q1 2022. How do they manage to have their returns stay flat?
Edit: Nevermind, it’s probably a matter of scale. The returns of T-Bills during that period were around -0.1% / -0.2%.
Edit 2: As it happens, while T-bills don’t distribute interests, T-bills funds may distribute dividends, which explains much of the decline I had noted.
Turns out this strategy is basically identical to the baseline scenario. Here a possible explanation from my side:
So, to sum up: It doesnt make sense split your investments during the month and try to time the day. But you might get a slight boost if you are willing to work harder to invest a bit more after a dip and
This of course would be different in a prolonged bear market, where you would save more money if you would keep this additional “miracle money” in a savings account. But to be fair, in this case you would save the most money if you would withdraw at the start of the bear run completely (but no one knows when, whether and how long that would take).
So, it was both a matter of scale and of me using a T-Bills fund ($BIL) assuming there would not be dividends while, actually, there were (distributed from the fund)…
I would challenge white coat investors to do something different, though (here comes the neurological mumbo jumbo). Every time you see the market go up, start cursing. Seriously! Cursing is usually associated with negative things in our lives, and the pathways involving the amygdala (fear center) and prefrontal/insular cortex (anger) will be activated and you will learn that every time the market goes up, it is bad. Also, while you curse, think of yourself eating Alpo in retirement and your kids going to community college despite a Harvard acceptance because you are too poor. Silly imagery, yes, but it will teach your brain that the market going up when you are nowhere near retirement is a BAD thing using System 1.
And then when the market goes down, say “Sweet!” and think of your kids (assuming you love your kids), wife, and family. Usually I do this when I see on my iPhone the market is down. I yell “SWEET!” and then scroll through photos of my family and kids. I am activating the limbic system and nucleus accumbens, the happiness and reward centers of our brains, in response to the market falling. So, you are adapting your System 1 to fear the market going up and to feel happy and want to invest when the market is going down.
By reading and partipating to this forum, you confirm you have read and agree with the disclaimer presented on http://www.mustachianpost.com/
En lisant et participant à ce forum, tu confirmes avoir lu et être d'accord avec l'avis de dégagement de responsabilité présenté sur http://www.mustachianpost.com/fr/
Durch das Lesen und die Teilnahme an diesem Forum bestätigst du, dass du den auf http://www.mustachianpost.com/de/ dargestellten Haftungsausschluss gelesen hast und damit einverstanden bist.
. More seriously, I am not sure how you have calculated the return and if it is correct. I don’t even know what is the right way to calculate it in this scenario.
