In headlines this week reads as "the market is falling," and it is not that. It is a single chain with a dated trigger, running through crude, from there to long-dated bonds, and from there into the private investor's portfolio.
Errors or inaccuracies are possible. Spotted something that looks wrong? Write to me and I will correct it.
A clarification before anything else, and it is worth reading.
This is a developing event. The geopolitical situation described here changes from day to day, and the figures in this piece are as of the close on 18 August 2026, except for the crude price, which includes 19 August as well. Every number here carries an explicit date so that it is possible to know when it was measured.
And causation in markets is inference, not proof. Nobody can prove that a price fell "because of" a particular piece of news. What can be done is to point to a dated event, show what moved afterwards and on what scale, and say explicitly where the line runs between the measurement and the explanation. That is what is done here.
And nothing here is a recommendation to take any action - not to buy, not to sell and not to hold. This is a piece that explains a mechanism.
The Week in Numbers
All figures are closing prices:
| 13 August | 17 August | 18 August | 13 to 18 | |
|---|---|---|---|---|
| S&P 500 (SPY) | 777.88 | 772.67 | 767.45 | -1.34% |
| Nasdaq 100 (QQQ) | 732.07 | 729.87 | 717.51 | -1.99% |
| ARKK - innovation basket | 82.59 | 81.75 | 79.15 | -4.17% |
| Long Treasuries (TLT) | 82.59 | 81.35 | 81.66 | -1.13% |
| Brent crude, $/barrel | 87.07 | 90.87 | 91.02 | +4.54% |
| 30-year yield | 5.21% | 5.31% | +10 bp |
Two rows in this table moved far more than the index, and they are the first two stops in the chain.
The Trigger: The Memorandum That Expired
On 17 June 2026, at the Palace of Versailles, during a G7 dinner, a 14-point memorandum of understanding was signed between the United States and Iran. Pakistan mediated, and the Iranian president signed it electronically from Tehran.
The memorandum was not a permanent agreement. It gave both sides 60 days to negotiate one, and within that window it set a ceasefire, the opening of the Strait of Hormuz to toll-free commercial passage, and an end to the naval blockade - alongside a framework for a permanent accord that included a $300 billion reconstruction fund.
Those sixty days ended on 17 August. No permanent agreement was signed.
And this is what turned that date into a market event rather than a purely diplomatic one.
The US president ruled out an extension of the truce, and said he does not see the war ending soon. The two sides do not even agree on what expired: Washington treats the date as the end of the ceasefire, while Tehran maintains that the memorandum concerned ending the war rather than a temporary truce.
And on the night between 17 and 18 August a cargo vessel in the Strait of Hormuz was struck by a projectile of unidentified origin.
The central point of dispute is not the nuclear file - it is the strait.
And Why the Strait of Hormuz Is the Junction
It is a narrow maritime passage through which, in normal times, roughly a quarter of the world's seaborne trade in crude oil and refined products passes, along with about a fifth of the trade in liquefied natural gas - on 2025 figures.
And it is effectively closed. It shut at the end of February, opened in early April, was declared open in mid-April and closed again a day later. The arrangement set out in the memorandum collapsed in early July following attacks on commercial vessels, and since then the passage has not functioned for routine commercial traffic.
Here is what that looks like in numbers: on 9 August a single ship crossed the strait. In normal times roughly 73 ships cross it per day.
The First Stop: Crude
| Brent, $/barrel | |
|---|---|
| 13 August | 87.07 |
| 14 August | 88.52 |
| 17 August | 90.87 |
| 18 August | 91.02 |
| 19 August | 91.86 |
Three consecutive rises, and then a fourth.
And here it is worth pausing on a methodological point, because it teaches something about reading data.
The sentence "crude has risen" and the sentence "crude is below its high" can both be true at the same moment - it all depends on the point you measure from and which series you are holding.
Here is the check that removes the doubt: Brent's late-July high was $90.74 on 29 July. The price on 18 August, $91.02, is above it. And on 19 August, $91.86, it is clearly above it.
So the claim that crude is "still below late-July levels" is not correct for these days. It is above them.
The practical conclusion for the reader: when somebody tells you a price "has not really risen," ask for the series and for the date of the last number in it. A series that stopped a week ago will give an answer that was true - and is not true today.
The Second Stop: Long-Dated Bonds
Here the move is larger than the one in equities, and this is the part most coverage skips.
| 17 August | Yield |
|---|---|
| Three months | 3.87% |
| Two years | 4.19% |
| Ten years | 4.72% |
| Thirty years | 5.31% |
5.31% is the highest yield on the 30-year US Treasury since June 2007 - that is, in about 19 years. And on 18 August it continued higher, to around 5.33%.
But the important number is not the level. It is the relative direction:
| From 13 to 17 August | Change |
|---|---|
| Three months | 0 basis points |
| Two years | +4 |
| Ten years | +9 |
| Thirty years | +10 |
And this is the distinction worth holding onto, because it says something entirely different from "rates are rising."
The short rate did not move at all. The three-month yield was 3.87% on 13 August, and 3.87% on 17 August.
When the short end freezes and the long end rises, that is not a change in expectations for the central bank's policy rate. Expectations of that kind show up at the short end and in the belly of the curve.
What shows up at the long end is the premium investors demand in order to hold long-dated debt - and into that premium go expectations of future inflation and the supply of debt.
And crude touches the first of those directly. More expensive energy over time means higher expected inflation, and that is priced at the long end of the curve, not the short end.
And the Second Leg at the Long End: The Deficit
Energy is not the only force at work there, and it would be wrong to hang everything on it.
The US federal deficit is running at roughly 6% of GDP, and the Congressional Budget Office raised its annual deficit forecast to $2.1 trillion - about $200 billion above its February estimate. Among the drivers cited: medical costs and the interest on the debt itself.
And in parallel, a wave of corporate debt issuance is competing for the same money.
And why this lands at the long end in particular.
A larger deficit means issuing government bonds in larger size. And when supply grows, buyers demand a higher yield in order to absorb it. That is the same premium discussed above, from the supply side rather than the inflation side.
And this is where the midterm elections enter, to be held on 3 November 2026, in which all 435 House seats and 35 Senate seats will be decided. The current House majority is 220 to 215 - the narrowest since 1930.
And the market-relevant point is not who wins. It is that the composition of Congress determines who legislates a budget, and fiscal policy is precisely the variable the long end prices. A divided government tends toward less major legislation, but also toward confrontations over spending caps and the debt ceiling.
This is not a political forecast and I do not make one. It is an observation about what sits on the market's calendar, and about why uncertainty over fiscal policy lives in long-dated bonds rather than short ones.
The Third Stop: And Here It Reaches Equities
The mechanism is simple, and worth knowing thoroughly.
The value of a share is the discounted value of future earnings. The higher the long-term risk-free yield, the less those future earnings are worth today.
And this does not affect all shares equally. A company whose earnings are mostly expected ten or twenty years out is far more sensitive to the long yield than a company earning today.
Which is why a ten basis point rise at the long end is a larger event, in pricing terms, than a half-percent fall in the index. The first changes the denominator for every asset; the second is one day's move.
And here is how that looked on 18 August:
| 18 August | Change on the day |
|---|---|
| S&P 500 | -0.68% |
| Nasdaq 100 | -1.69% |
| ARKK | -3.18% |
| Long Treasuries (TLT) | +0.38% |
ARKK fell 4.7 times as much as the broad index on that same day.
And Why It Feels Like Everything Is Falling
Here is the gap for which this piece was written.
The broad index is down 1.34% from its 13 August peak. That is a pullback, not a crash. And yet for a large share of private investors the experience is of a far harsher week - and they are not mistaken.
The reason is that the average private investor's portfolio is not the index.
From the 13 August peak through 18 August:
| From the peak | |
|---|---|
| S&P 500 | -1.34% |
| Nasdaq 100 | -1.99% |
| ARKK | -4.17% |
And because the private investor is usually concentrated in growth, in technology and in thematic positions rather than in a broad index, the number showing up in the portfolio is far closer to the third row than to the first.
In other words, the headline reports the movement of the first row, and the experience is set by the third.
And this also explains another sensation: somebody holding several different growth names sees all of them red on the same day, and concludes that "everything is falling." In practice they are not falling separately - they are all exposed to the same variable, the long yield. Diversification across five growth companies is not diversification against that risk.
And here is the figure that sharpens this more than any other:
| End of March to 18 August | |
|---|---|
| S&P 500 | +18.01% |
| Nasdaq 100 | +24.31% |
| ARKK | +17.10% |
At the 17 August close, ARKK was still ahead of the broad index - +20.95% since the end of March against +18.81%.
One trading day later it is below it.
הזווית שלי
דעה אישית של אילן אברמוב - לא ייעוץ ולא המלצה
What I take from this week is mainly order, not a conclusion.
When an index falls a little over one percent, the question "why" almost always produces an answer - because there is always some piece of news to point at. What is rarer is a week in which the trigger carries a precise date, and can be followed stop by stop. A memorandum that expired on 17 August, crude that rose four sessions in a row, a long yield at a 19-year high, and a growth basket down 4.7 times the index. That ordering is coherent, and that is not always the case.
And what I think deserves attention is the fiscal leg rather than the geopolitical one. Geopolitical tension is by its nature the kind of event that can reverse on a single announcement. A deficit of 6% of GDP does not reverse on an announcement. So if the long yield stays at these levels even after crude settles - that will be the sign that the story was never Iran in the first place, but supply.
And the part I consider most important in this piece is the last one. The gap between -1.34% in the index and -4.17% in the growth basket is the gap between what the news reports and what people feel. And somebody who does not know they are holding an asset with two or three times the amplitude of the index finds out at exactly the worst moment to find out - inside the decline.
This understanding is not an argument against growth. It is an argument for knowing what you hold before the market reminds you. An asset's historical volatility is available information, and it is known in advance; the feeling inside the decline is the part that cannot be forecast.
(It is important to stress: this is my personal opinion only, and nothing herein constitutes a recommendation to take any action.)
