When is the right time to buy
Buying at once, or spread over months?
A family has USD 10m to put into the market. Every trading day of the last ten years was tested as a start day: buying at once against equal monthly pieces over 3 to 24 months, and what bad timing cost.
VOO and QQQ · USD 10m · daily closes 27 Oct 2016 to 27 Oct 2026, dividends reinvested, waiting cash in T-bills (BIL)
The headline
Buying at once won about eight times in ten
Spreading over 12 months against buying at once, compared on the day of the last purchase, when both are fully invested. 2,284 start days per fund.
82% · 83%
Start days on which the lump sum finished ahead, VOO · QQQ
+5.9% · +9.1%
Typical gap in favour of the lump sum: USD 590k · 910k on 10m
−34% · −25%
Worst point for the 10m on VOO, lump sum · spread
708 · 429 days
Longest run below 10m on VOO, lump sum · spread
By start day
Spreading paid only in the weeks before a top
Each point is a start day: the lump sum’s value minus the 12-month plan’s value at the last purchase, as a share of the 10m. Navy is the lump sum ahead. The bronze patches are the only times spreading paid: the weeks before the September 2018, February 2020 and January 2022 tops.
VOO
QQQ: the same picture, with bigger swings both ways
Length of the spread
The longer the spread, the bigger the cost
The longer the money is spread, the more often and the more it costs. What it buys is a shallower worst point and a shorter time underwater.
| Fund | Spread over | Lump sum ahead | Typical gap | Worst point, lump / spread | Longest underwater, days |
|---|---|---|---|---|---|
| VOO | 3 months | 72% | +1.5% | −34 / −32% | 708 / 561 |
| VOO | 6 months | 76% | +3.1% | −34 / −29% | 708 / 464 |
| VOO | 12 months | 82% | +5.9% | −34 / −25% | 708 / 429 |
| VOO | 18 months | 87% | +8.4% | −34 / −21% | 708 / 402 |
| VOO | 24 months | 88% | +11.6% | −34 / −19% | 708 / 302 |
| QQQ | 3 months | 69% | +1.9% | −35 / −34% | 715 / 702 |
| QQQ | 6 months | 78% | +4.2% | −35 / −31% | 715 / 554 |
| QQQ | 12 months | 83% | +9.1% | −35 / −28% | 715 / 508 |
| QQQ | 18 months | 87% | +12.8% | −35 / −25% | 715 / 484 |
| QQQ | 24 months | 86% | +16.2% | −35 / −21% | 715 / 448 |
Bad timing
USD 10m put in at the top of each decline
The worst case: 10m invested on the top day before every decline of 8% or more. Value today (27 Oct 2026) and the worst point on the way. Highlighted cells are where a spreading plan finished ahead of the lump sum.
VOO
| Top | Fell | Back at the top | All at once | 6 months | 12 months | 24 months | At the low instead |
|---|---|---|---|---|---|---|---|
| 26 Jan 2018 | −10% | 180 days | 30.8m / −19% | 32.4m / −15% | 32.4m / −15% | 31.3m / −18% | 34.3m |
| 20 Sep 2018 | −19% | 204 days | 29.8m / −22% | 32.5m / −15% | 31.3m / −18% | 30.0m / −15% | 37.1m |
| 19 Feb 2020 | −34% | 173 days | 25.1m / −34% | 28.8m / −7% | 26.4m / −3% | 22.9m / −1% | 38.0m |
| 2 Sep 2020 | −9% | 70 days | 23.5m / −9% | 23.5m / −2% | 21.7m / −5% | 20.3m / −11% | 26.0m |
| 3 Jan 2022 | −25% | 709 days | 17.2m / −25% | 18.7m / −18% | 19.9m / −11% | 19.8m / −5% | 22.8m |
| 16 Jul 2024 | −8% | 65 days | 14.0m / −11% | 13.9m / −12% | 13.9m / −9% | 13.0m / −3% | 15.3m |
| 19 Feb 2025 | −19% | 127 days | 12.8m / −19% | 13.6m / −5% | 12.8m / −2% | not finished | 15.8m |
| 27 Jan 2026 | −9% | 77 days | 11.2m / −9% | 11.1m / −2% | not finished | not finished | 12.2m |
QQQ
| Top | Fell | Back at the top | All at once | 6 months | 12 months | 24 months | At the low instead |
|---|---|---|---|---|---|---|---|
| 29 Aug 2018 | −23% | 226 days | 40.7m / −23% | 45.1m / −12% | 43.4m / −5% | 39.9m / −5% | 52.8m |
| 3 May 2019 | −11% | 61 days | 39.6m / −11% | 41.2m / −7% | 39.2m / −11% | 33.2m / −5% | 44.4m |
| 19 Feb 2020 | −29% | 105 days | 31.7m / −29% | 33.6m / −5% | 29.7m / −2% | 25.4m / −10% | 44.4m |
| 2 Sep 2020 | −13% | 90 days | 24.7m / −13% | 25.2m / −11% | 23.5m / −17% | 22.4m / −21% | 28.3m |
| 12 Feb 2021 | −11% | 56 days | 22.2m / −22% | 22.3m / −21% | 20.9m / −26% | 22.9m / −19% | 24.9m |
| 27 Dec 2021 | −35% | 716 days | 18.4m / −35% | 21.5m / −24% | 23.6m / −17% | 23.3m / −8% | 28.4m |
| 10 Jul 2024 | −14% | 119 days | 14.5m / −17% | 15.1m / −14% | 15.1m / −10% | 13.8m / −3% | 16.8m |
| 19 Feb 2025 | −23% | 125 days | 13.5m / −23% | 14.4m / −6% | 13.4m / −3% | not finished | 17.5m |
| 29 Oct 2025 | −12% | 168 days | 11.4m / −12% | 11.9m / −8% | 11.2m / −3% | not finished | 12.9m |
Even the worst entries of the decade are well above 10m today. Spreading softened the worst point every time; it finished ahead only after the three big tops.
Time in the market
Whenever you bought, the time in the market did the work
USD 10m put in at once on each day of the last ten years, valued at the close of 27 Oct 2026, in million USD. Only 20 VOO and 44 QQQ start days out of 2,513 are below 10m today.
The year mattered more than the day
The worst day of 2019 is worth 26.3m on VOO today. The best day of 2021: 22.6m.
The best entries were the lows: 24 Dec 2018, 23 Mar 2020, 12 Oct 2022 and 8 Apr 2025.
Even the January 2022 top, the worst day before this year, is 17.2m on VOO and 18.4m on QQQ today.
What to expect after buying
A dip was normal, a long wait was rare
USD 10m at once: share of start days of the last ten years with a full year after them (2,262 per fund).
The long waits had one source
Every start day underwater for over a year came from early 2021 to mid 2022.
That was 42% of VOO’s 2021 start days and 65% of QQQ’s, 25 to 27% of 2022’s, and none from any other year.
The worst was the January 2022 top: 708 days. Nothing lasted two years. 2000 to 2002 and 2008, outside the data, did.
Seven ways in
Which way in kept the wait shortest
Each rule against putting it all in at once, VOO / QQQ, on the same 2,262 start days. Median cost is what the rule gave up against the lump sum. Rules waiting for a dip buy the rest after 12 months if the dip never comes.
| Way in | Median cost | Longest underwater, days | Underwater more than a year |
|---|---|---|---|
| All at once | – | 708 / 715 | 8 / 10% |
| 12 monthly pieces | 5.7 / 8.4% | 429 / 508 | 5 / 11% |
| Half at once, half in 12 pieces | 2.8 / 4.2% | 525 / 529 | 7 / 12% |
| Half at once, half 10% below the start price | 4.4 / 6.0% | 450 / 598 | 2 / 11% |
| A third at once, thirds at −10% and −20% | 5.9 / 8.1% | 412 / 508 | 1 / 8% |
| Out to T-bills below the 10-month average | 26.5 / 21.1% | 893 / 486 | 20 / 4% |
The two funds favoured different rules
VOO: a third now and the rest after 10% and 20% falls worked best. Only 1 in 100 start days stayed underwater for more than a year, against 8 in 100 for buying all at once.
QQQ: the same rule barely helped (10 to 8 in 100). Moving to T-bills below the 10-month average did best, 4 in 100, but on the median start day it ended 21% behind simply holding QQQ.
Dip rules need a dip: a 10% fall came within a year on only 40% of start days.
What the rules show
Recovery only got shorter when there was still money left to buy lower
01
Nothing removed the 2021–22 wait.
A recovery only got shorter when cash was still there to buy lower, so the only lever is money kept back for declines.
02
The reserve bought on declines did the most for VOO.
Thirds at once, at −10% and at −20% cut the start days with a year underwater from 8% to 1% and the worst wait from 708 to 412 days, for a median 6% of the money.
03
It only works when the dip comes.
A 10% fall below the start price came within a year on only 40% of start days. The other 60% bought the rest a year later at a higher price, which is where the 6% goes.
04
Spreading cut the depth, not QQQ’s waits.
The 12-month plan kept buying through 2021–22 and sat underwater almost as long.
05
The trader’s exit was the worst way in.
Out of the market 17% of the time, with 20 switches on VOO, it missed the rebounds of 2019, 2020, 2023 and 2025. It never beat holding VOO and cost 21% on QQQ, where it did cut the long waits from 10% to 4% of start days.
What ten years say
Four conclusions
- Invest when the money arrives. Ahead on 7 to 9 start days out of 10; spreading cost a typical 3 to 12% of the money.
- Spreading is insurance against a top. It paid only within weeks of September 2018, February 2020 and January 2022, by 6 to 17% over 6 months.
- If a long wait would hurt, keep a reserve for declines. A third at once and thirds at −10% / −20% cut VOO’s worst wait by 40%, for about 6%.
- Do not time exits. The 10-month trend rule cost 21 to 27% and never beat holding VOO.
Past ten years only, with one long bear market in the sample. Not investment advice.
Important information
This page is for general information only and is directed at professional investors and family office principals. It is not an offer, solicitation or recommendation to buy or sell any security or financial product, and it is not investment, legal or tax advice.
Results are historical back-tests of VOO and QQQ over 27 October 2016 to 27 October 2026, using dividend-adjusted daily closes with waiting cash held in T-bills (BIL). Spreading plans buy 10m / N on the start day and each monthly anniversary and are compared with the lump sum at the last purchase. The period contains one long decline. Thresholds and rules were chosen, not optimised, and outcomes depend heavily on the start date. Past performance is not a reliable indicator of future results, and the value of investments can fall as well as rise.
Any services are provided by Goltsblat Capital only under a written engagement and subject to applicable laws and regulations.