Should you buy the ASX 200 at all-time highs?
Records highs aren't sell signals. 26 years of ASX data show buying at a peak often pays off better a year later than buying any random day.
Source: Shutterstock
KEY POINTS
- Buying the ASX 200 at record highs has historically led to slightly weaker returns in the first few months but noticeably better returns a year out compared to buying on any random day.
- The pattern holds up even after decluttering the data, via a pullback filter or excluding pre-GFC and pre-pandemic highs, meaning a record high is, at worst, a slightly bumpier entry and, at best, a much stronger one.
- The current all-time high is backed by solid fundamentals, including forecast earnings growth of roughly 12-13% in FY26/FY27, easing oil prices and rate expectations, reset positioning after the July tech selloff, and a strong US earnings season.
The S&P/ASX 200 cleared its previous 2 March record to close at its fifth all-time high of the year. It feels like a long time since "ASX 200" and "record high" have shared a sentence, after five exhausting months of US–Iran war headlines, inflation and fuel-price concerns, the end of the housing supercycle and much more.
For many investors, the instinct at an all-time high is to sell, as a correction or pullback can feel almost inevitable.
That's what we're here to answer today. Is it actually a good idea to buy at all-time highs?
Buying the ASX 200 at all-time highs
For starters, the technically correct benchmark is the S&P/ASX 200 Total Return Index. But let's be honest, no one checks that one, so we'll use the price index like everyone else and keep in mind that dividends are quietly doing some of the work.
Since 2000, the S&P/ASX 200 has recorded 288 all-time highs, so roughly one every 23 trading days. They tend to cluster tightly, especially across 2004–07, 2019–20 and 2024–26, with an almost 12-year gap from 2008 to 2018 where zero record highs were made.
Here's how the market performed after an all-time high.
1-Week | 1-Month | 3-Month | 6-Month | 12-Month | |
|---|---|---|---|---|---|
Average | -0.27% | -0.12% | 0.69% | 2.60% | 6.19% |
Median | 0.11% | 0.38% | 0.99% | 3.09% | 8.03% |
% Positive | 55% | 58% | 56% | 68% | 68% |
S&P/ASX 200 forward returns after an all-time high, January 2000 – August 2026. Source: Market Index
The returns and % positive look fairly decent at the 12-month mark, but how does that compare to buying on any other day? The "base" figures below show the forward return from any trading day across the same period.
1-Week | 1-Month | 3-Month | 6-Month | 12-Month | |
|---|---|---|---|---|---|
Base average | 0.10% | 0.43% | 1.24% | 2.51% | 5.02% |
Base median | 0.24% | 0.82% | 1.87% | 3.39% | 5.81% |
ATH vs. Base (median) | -0.13% | -0.44% | -0.88% | -0.30% | 2.22% |
S&P/ASX 200 forward returns for all trading days between January 2000 and August 2026 | Source: Market Index
So buying at all-time highs has, on average, led to slightly weaker returns in the short term but better returns a year out, and by a meaningful margin.
Now, you can chop this data a few ways. We've counted every single all-time high, but they tend to arrive in clusters. So what happens if we add a rule where the index must first fall at least 1% below the prior peak before a new high "counts"? This declutters the count to just 64 highs.
1-Week | 1-Month | 3-Month | 6-Month | 12-Month | |
|---|---|---|---|---|---|
Average | -0.55% | -1.03% | -0.78% | 0.99% | 2.47% |
Median | -0.24% | 0.08% | -0.18% | 2.03% | 3.84% |
% Positive | 46% | 52% | 48% | 61% | 62% |
S&P/ASX 200 forward returns after an all-time high, January 2000 – August 2026. A 1% dip resets the counter. Source: Market Index
The average returns are now much weaker, as decluttering strips out the mid-trend highs and leaves a list weighted towards records made just before a pullback.
But what about the cluster of all-time highs right before the GFC and pandemic? If we go back and count every high, and do the opposite (exclude the two pre-crash clusters, 22 in total). Now, the data turns decisively more bullish.
1-Week | 1-Month | 3-Month | 6-Month | 12-Month | |
|---|---|---|---|---|---|
Average | −0.24% | 0.12% | 1.92% | 4.18% | 8.22% |
Median | 0.11% | 0.34% | 1.57% | 3.79% | 10.95% |
% Positive | 55% | 58% | 60% | 74% | 74% |
S&P/ASX 200 forward returns after an all-time high between January 2000 and August 2026, excluding the 22 ATHs prior to the GFC and pandemic | Source: Market Index
The bottom line
Slice it however you like, count every high, one per 1% dip, or strip out the pre-crash tops, and the data tells a similar story. Buying the ASX 200 at a fresh all-time high is, at worst, a slightly bumpier version of buying on any other day, and a touch better a year out. What it plainly isn't is a sell signal.
The market is breaking out after a year of sideways trade, and consensus now points to roughly 12% earnings growth in FY26 and 13% in FY27, led by the Materials sector. The macro backdrop is accommodative too, with oil prices easing, yields and interest rate expectations are drifting lower, a July selloff across semis and AI-related names reset stretched positioning, and US earnings season has been remarkably strong. Records highs tend to appear in clusters, and if this is the start of another one, history says there's little reason to stay out of it.
