Daniel sent us this one — he's been watching the Bank of Israel's intervention on June fourth, where they dropped eight hundred and one million dollars in a single day buying shekels to offset what they called speculative volatility. The question is: what volume of institutional money does it actually take to move a major exchange rate, and who are the players placing those bets? Because eight hundred million sounds like a lot. But in the global foreign exchange market, the question is whether it's a cannonball or a pebble.
Right, and that's exactly the tension. The global FX market turns over about seven and a half trillion dollars every single day. So eight hundred and one million is roughly one hundredth of one percent of daily global volume. It's a rounding error. But here's the thing — the shekel is not the euro. USD ILS daily spot turnover, according to the most recent BIS triennial survey data, is roughly five to seven billion dollars a day. Now that eight hundred and one million is somewhere between twelve and fifteen percent of daily volume in that specific pair. That's not a rounding error anymore. That's a meaningful presence.
It's the difference between shouting in a stadium and shouting in a library.
In euro dollar, which does over seven hundred billion a day in spot alone, eight hundred million is a quiet Tuesday afternoon for a single large asset manager rebalancing a portfolio. The market absorbs it without blinking. In shekel dollar, that same amount is an event. And the Bank of Israel knows this. That's why they intervened — because in a thin market, a determined institutional player can cause outsized moves with relatively modest amounts of capital.
They specifically called out speculative investments. Which is interesting language for a central bank. They're not saying the shekel is mispriced. They're saying someone is gaming it.
Yes, and that distinction matters. When a central bank says it's countering speculative volatility, it's drawing a line between what it considers legitimate two way flow and what it sees as predatory or destabilizing positioning. So let's unpack what speculative actually means here. It's not some hedge fund manager in a dark room twirling a mustache. It's typically one of three things. One, carry trades — borrowing in a low yield currency like dollars or yen and parking the money in shekel denominated bonds that yield more. Israel's interest rates have been relatively attractive for a while now. Two, algorithmic momentum strategies that detect a trend and pile on. If the shekel has been strengthening for six weeks, the algos see that and go long. Three, leveraged macro funds taking directional bets based on their view of Israeli economic fundamentals or regional stability.
It's fast money versus slow money. The fast money sees a trend, rides it, and gets out. The central bank is trying to say — no, the trend is not the truth.
The problem with fast money in a thin market is that it becomes the market. If daily volume is five to seven billion, and a handful of macro funds each put on two hundred to three hundred million dollar positions over the course of a week, they can create the very trend their algorithms are detecting. It's a feedback loop. The shekel strengthens, the momentum algos buy more shekels, the strengthening accelerates, corporate treasuries panic and start hedging more aggressively, and suddenly you've got a currency that's up five percent in a quarter for no fundamental reason.
Which brings us to the core question. How much money does it actually take to move a rate like USD ILS by, say, one or two percent? Is there a threshold?
There's academic work on this. Sarno and Taylor did the foundational research back in two thousand one, and Fratzscher updated it more recently. The rough consensus is that to move a mid sized currency pair by one to two percent, you need order flow equivalent to about five to ten percent of daily turnover. For USD ILS, that's somewhere between two hundred fifty million and seven hundred million dollars. The Bank of Israel's eight hundred and one million is at the upper end of that band. So they're not messing around — they came in with enough firepower to actually move the needle.
That's intraday. That's the initial shove. What about keeping it there?
That's the harder question. Because moving a rate intraday is one thing. Sustaining that move requires changing the positioning of the big institutional players. And that's where the concept of order flow versus positioning comes in. Order flow is the actual transaction — someone buying or selling shekels right now. Positioning is the accumulated net exposure of the major players. A central bank can dominate order flow for a day. It cannot dictate positioning for a quarter unless the fundamentals eventually validate the move.
It's like a parent grabbing the steering wheel. You can correct the swerve. But if the kid keeps yanking it left, eventually you're just wrestling.
Sometimes you lose. Let me give you a comparison that puts this in perspective. In twenty twenty two, the Bank of Japan intervened to defend the yen. They spent twenty billion dollars in a single day. That's twenty five times what the Bank of Israel just did. And it moved USD JPY by five percent in one day. But here's the context — daily turnover in dollar yen is about one hundred and twenty billion. So that twenty billion was roughly seventeen percent of daily volume. Similar proportion to the Bank of Israel's move, actually. The difference is scale, not ratio.
The ratio is what matters. Twelve to fifteen percent of daily volume is the central bank saying, we are the market today.
For today, yes. And Japan did that repeatedly in twenty twenty two. Three interventions over several weeks, totaling over sixty billion dollars. It worked temporarily. The yen strengthened. But the fundamental driver — the interest rate differential between the US and Japan — didn't change until the Fed started signaling rate cuts much later. So the market kept testing them.
Which brings up the Swiss National Bank, which is the cautionary tale everyone in FX circles knows. They spent years defending a cap on the euro franc exchange rate at one point two. They accumulated reserves worth eighty percent of Swiss GDP buying euros to hold that line. Eighty percent of GDP. And then in twenty fifteen they just... And the franc appreciated thirty percent in minutes.
That moment rewired how markets think about central bank credibility. The SNB had unlimited firepower in theory — they could print Swiss francs forever to buy euros. But the political and economic cost of accumulating that many reserves became unbearable. The balance sheet risk was enormous. And when they blinked, the market exacted a brutal price. The lesson for every central bank since then is that intervention without eventual fundamental alignment is just renting time.
The Bank of Israel has its own version of this story. October twenty twenty three, right after the attacks, they sold thirty billion dollars of reserves in a single month to defend the shekel. That was thirty percent of their total reserves.
In one month. And it worked — temporarily. The shekel stabilized for a few weeks. But over the following quarter, it still depreciated ten percent. Because the fundamental reality had shifted. The country was at war. No amount of intervention was going to change that narrative. The market understood that the Bank of Israel was buying time, not changing the trajectory.
If the fundamentals are against you, intervention is basically a very expensive pause button.
And that's the context for understanding the June fourth intervention. This time, the fundamentals are actually quite strong. Israel's economy has been resilient. The tech sector is still attracting capital. Interest rates are relatively high, which attracts carry trade inflows. The problem is the opposite of twenty twenty three — the shekel is too strong, not too weak. It's been appreciating rapidly, and that's hurting exporters.
Who are the actors on the other side? If the Bank of Israel is buying dollars to weaken the shekel, who's selling them those dollars?
This is where it gets interesting. Let me break down the four categories of institutional players who actually move exchange rates. First, global macro hedge funds. Bridgewater, Brevan Howard, Renaissance Technologies. Bridgewater alone manages over a hundred billion dollars. These funds run systematic FX strategies. They can allocate five to ten percent of their portfolio to a single currency bet. For a fund with fifty billion under management, that's two and a half to five billion dollars. They don't put that on all at once — they use algorithms to slice it into pieces over days or weeks — but the cumulative position can be enormous for a thin market like the shekel.
These are the people the Bank of Israel is implicitly wagging a finger at when they say speculative investments.
Some of them, yes. But here's where the misconception comes in. Not all speculative flow is hedge funds gambling. The second category is asset managers — BlackRock, Vanguard, State Street. These are the supposedly boring, long only institutions that manage pensions and retirement accounts. But they have to hedge currency exposure on their international portfolios. If Vanguard's total international bond fund decides to increase its allocation to Israeli bonds by half a percent, that could mean a five hundred million dollar flow into shekels over a quarter. That's not speculation in the pejorative sense — it's prudent portfolio management. But it has exactly the same effect on the exchange rate as a hedge fund placing a directional bet.
The distinction between speculative and legitimate is mostly about vibes.
It's about intent and time horizon, but the market impact is identical. A dollar of shekel buying from BlackRock rebalancing a pension fund moves the rate exactly as much as a dollar of shekel buying from a macro hedge fund running a carry trade. The central bank doesn't get to filter which dollars it absorbs.
That's a genuinely important point that most coverage misses. Everyone wants a villain, and the hedge fund manager in Greenwich makes a better villain than the pension fund administrator in Valley Forge.
And the third category makes this even murkier — corporate treasuries. Apple, Toyota, any multinational with significant revenue exposure. These companies have treasury departments that manage billions in currency hedging. If Apple expects the shekel to keep strengthening and they have significant Israeli revenue, their treasury team might accelerate hedging, buying shekels forward to lock in favorable rates. That's not speculation — it's risk management. But multiply that by dozens of multinationals and you've got a wall of institutional flow that a central bank has to swim against.
The fourth category?
Algorithmic high frequency trading firms. XTX Markets, Citadel Securities, Jump Trading. These firms provide liquidity — they're the ones making markets and earning the spread. But their algorithms also detect momentum. If the shekel has been trending stronger for six weeks, their models will be net long, and they'll amplify every uptick. They're not taking multi week directional views. They're operating on microsecond timeframes. But the aggregate effect of their positioning can be substantial.
The Bank of Israel is essentially playing a multiplayer game where it can see some of the players but not all of them, and the ones it can't see might actually be the largest.
The ones it can see might not even be the problem. A corporate treasurer hedging legitimate exposure isn't a speculator. But if enough of them hedge at the same time because the shekel is strengthening, they create the very volatility the central bank is trying to fight. It's reflexive. The intervention itself can trigger more hedging if corporates think the central bank is signaling that the shekel is overvalued.
Let's talk about the volume threshold for each of these actors. You mentioned a single macro fund can place a two hundred to five hundred million dollar FX trade without moving the market if they use algorithms. What does moving the market actually look like at different sizes?
In USD ILS, a two hundred million dollar trade executed carefully over a few hours might move the rate ten or twenty basis points — that's a tenth to a fifth of a percent. Noticeable to a trader watching the screen, but not headline worthy. A five hundred million dollar trade, especially if it's one sided and the market is thin at that moment, could move the rate half a percent. And a two billion dollar trade — which is the kind of flow a major asset manager might generate when rebalancing a large international portfolio — could cause a one to two percent intraday move. That's when you get the central bank's attention.
The Bank of Israel's eight hundred and one million dollar intervention is roughly equivalent to saying: whatever you just did, we're undoing it.
They're not just matching the flow. They're trying to overshoot enough to change the narrative. Because here's the thing about central bank intervention — it's as much about signaling as it is about raw volume. When the Bank of Israel steps in and buys eight hundred million dollars worth of shekels, they're not just absorbing sell orders. They're sending a message to every macro fund, every asset manager, and every corporate treasurer: we think the shekel is strong enough, and we have two hundred billion dollars in reserves. Want to test us?
Two hundred billion in reserves is the number to keep in mind. Because that's the ultimate backstop. The market knows the Bank of Israel can keep doing this for a while.
Two hundred and six billion as of the most recent data, yes. So an eight hundred million dollar intervention is less than half a percent of reserves. They could do this every week for a year and still have plenty of ammunition. That's a very different credibility profile than, say, Turkey.
The other cautionary tale.
Turkey spent over a hundred billion dollars defending the lira in twenty twenty three and twenty twenty four. And the lira kept falling. Because the fundamental problem wasn't speculative attack — it was monetary policy. The central bank was cutting rates into high inflation. Every trader on the planet could see the contradiction. So the market — domestic investors converting savings to dollars, foreign hedge funds shorting the lira — was too large and too motivated. No amount of intervention was going to work until the policy framework changed.
Credibility is the multiplier. With it, eight hundred million can do the work of eight billion. Without it, a hundred billion buys you nothing.
And the Bank of Israel has credibility. They've managed the shekel through multiple wars, political crises, and global shocks. When they intervene, the market pays attention. But credibility is also a depletable resource. Every intervention that doesn't eventually align with fundamentals erodes it a little bit. The SNB learned that the hard way.
Let's dig into the mechanics of how these institutional trades actually execute. You mentioned algorithms slicing up orders. What does that look like in practice, and why does it matter for the intervention question?
If you're a macro hedge fund and you want to buy five hundred million dollars worth of shekels, you don't just call a bank and say, give me five hundred million dollars of shekels at market. That would cause the rate to spike instantly, and you'd get a terrible price. Instead, you use an execution algorithm. The algo breaks your order into hundreds or thousands of smaller pieces — maybe five million each — and feeds them into the market over several hours or even days, using patterns designed to look like random noise rather than a large directional bet.
It's like trying to empty a swimming pool with a teaspoon, but you've got ten thousand teaspoons working simultaneously.
The algos are smart enough to vary the timing and size to avoid detection. Some use volume weighted average price, or VWAP, strategies that spread the order in proportion to normal trading volume throughout the day. Others use implementation shortfall algorithms that try to minimize the difference between the decision price and the execution price. The point is, a sophisticated institutional player can put on a very large position without ever showing their hand.
Which means the central bank might not even know who it's fighting until well after the fact.
The Bank of Israel sees the aggregate order flow in real time — they know someone is buying shekels. But they don't know if it's Bridgewater putting on a new carry trade, or BlackRock rebalancing, or Apple hedging next quarter's Israeli revenue. They just see the buying pressure. And they have to decide: is this flow that will reverse naturally, or is this the beginning of a trend that will feed on itself?
That's where the judgment call comes in. And the June fourth intervention suggests they decided it was the latter.
And the specific language — irregular volatility from speculative positions — tells us they saw patterns in the order flow that looked like momentum algorithms or concentrated hedge fund positioning rather than diverse, fundamentals driven buying. Central banks have access to granular transaction data that we don't see. They can distinguish between a market where fifty different institutions are each buying ten million dollars of shekels for different reasons, and a market where three funds are each buying a hundred and fifty million through the same algo pattern.
What does success look like for this intervention? How do we know if it worked?
There are a few metrics. The most immediate is whether the shekel stopped appreciating. If the rate stabilizes or reverses modestly, the intervention achieved its tactical objective. The second is whether volatility declines — the Bank of Israel specifically cited volatility as the problem, not the level. The third, and this is harder to measure, is whether the speculative positioning unwinds. If the macro funds decide the shekel has peaked and start taking profits, the intervention has succeeded strategically.
The shekel has stabilized since June fourth, but as you said earlier, we don't know if that's because of the intervention or because the macro environment shifted.
That's the perennial problem in evaluating central bank interventions. You can never run the counterfactual. We don't know what the shekel would have done without the intervention. Maybe the speculative flow was about to exhaust itself anyway. Maybe the intervention was the only thing preventing a further three percent appreciation. We'll never know with certainty.
Which is why central banks tend to get more credit than they deserve when interventions appear to work, and more blame than they deserve when they fail.
The market is constantly updating its assessment of central bank credibility. Every intervention is a data point. If the Bank of Israel intervenes and the shekel stabilizes, the market assigns a higher probability that future interventions will also work. That makes future interventions more effective because speculators are less willing to bet against a central bank with a track record. It's a virtuous cycle. Until it isn't.
Let's talk about the shekel specifically as a market. You mentioned it's thin relative to G10 pairs. Just how thin are we talking?
USD ILS daily spot turnover of five to seven billion compares to euro dollar at over seven hundred billion. That's a factor of one hundred. The shekel is also thin compared to other emerging market currencies. Dollar peso does about fifteen to twenty billion a day. Dollar rand does about ten to twelve billion. So the shekel is on the smaller end even within emerging markets. That thinness is both a vulnerability and, paradoxically, a source of strength for the central bank.
In a thin market, a determined central bank has more influence. Eight hundred million dollars is twelve to fifteen percent of daily volume in shekel dollar. To achieve the same proportional impact in euro dollar, the ECB would need to deploy about a hundred billion dollars in a single day. That's not happening. So the Bank of Israel's relative firepower is actually much greater than the ECB's or the Fed's, even though their absolute reserves are much smaller. The thinness that makes the shekel vulnerable to speculative attack also makes it more defensible.
It's a double edged sword. The same thinness that lets a hedge fund move the rate also lets the central bank move it back.
And the Bank of Israel has demonstrated repeatedly that it understands this asymmetry and is willing to exploit it. The October twenty twenty three intervention — thirty billion in a month — was a statement of intent as much as a financial operation. They were saying: we know this market is thin, we know we can move it, and we will.
Let's circle back to the specific actors. You mentioned Brevan Howard and Bridgewater. Are there funds that specialize specifically in Israeli markets?
There are funds with dedicated Israel mandates, yes. But more commonly, Israel exposure comes through broader emerging market or global macro mandates. A fund like Brevan Howard, which runs about twenty billion plus in macro strategies, might have a dedicated EM FX pod that trades shekel as part of a broader basket. They're not betting on Israel specifically — they're betting on a carry trade or a momentum signal that happens to include the shekel. The distinction matters because it means the flow can reverse quickly if the signal changes, regardless of what's happening in the Israeli economy.
Which makes it even harder for the central bank to predict or manage. They're not just responding to views on Israel. They're responding to global risk appetite, Fed policy expectations, and algorithmic signals that have nothing to do with Tel Aviv.
And this is the challenge for any small open economy with a floating currency. Your exchange rate is determined not just by your own fundamentals, but by global capital flows that are increasingly automated and increasingly concentrated in a handful of large institutional players. The top ten global asset managers control over forty trillion dollars in assets. The top ten hedge funds manage over five hundred billion. A tiny shift in their portfolio allocations can swamp a small currency market.
The Bank of Israel isn't really fighting speculators in the old fashioned sense. It's fighting the gravitational pull of global capital allocation.
That's a much harder fight. Because you can't outgun the global financial system. You can only influence the terms on which it interacts with your currency. Which brings me to something I think is underappreciated about the Bank of Israel's approach. They're not trying to peg the shekel or target a specific level. They're trying to manage the pace of adjustment and reduce volatility. It's a subtle but important distinction.
They're not saying the shekel should be at three point six or three point seven. They're saying: whatever the right level is, let's get there without whiplash.
And that's a more defensible position. If you target a specific level, the market will eventually test you, and you'll either win or lose in a binary way. If you target volatility, you can claim success as long as the moves are orderly, regardless of direction. It's a smarter mandate.
Although it does raise the question of whether the market eventually figures out that you're not defending a level and just trades through your smoothing operations.
That's the cat and mouse game. And this is where I think the next frontier is interesting. Central banks are starting to develop their own algorithmic execution capabilities to counter the HFT firms and macro funds. Instead of a human trader at the Bank of Israel calling up a bank and placing an eight hundred million dollar order, imagine an algo that detects unusual momentum in shekel trading and automatically places countervailing orders in small slices, mimicking the same slicing techniques the speculators use.
The central bank becomes a stealth market participant rather than a visible hand.
Some already are. The Bank for International Settlements has published papers on this. The idea is that if the market knows the central bank is intervening, it front runs the intervention. But if the intervention looks like normal order flow, it achieves the same volume impact without the signaling effect — or rather, with a different, more ambiguous signal. The market knows the central bank might be in there, but it can't be sure.
Which is its own kind of psychological warfare.
Central banks are increasingly comfortable with that. They've learned that transparency is not always a virtue in FX markets. Sometimes you want the market to wonder.
Let's bring this back to something practical for anyone listening who actually manages currency exposure or trades FX. What should they be watching to understand whether the Bank of Israel's intervention is working and what comes next?
Two data points are especially useful. First, the Bank of Israel's reserve data, which is published monthly. A sudden drop in foreign currency reserves signals intervention — they sold dollars to buy shekels. The June data will show an uptick in reserves because they were buying dollars this time, not selling. But the principle is the same. Watch the month over month change. If reserves jump by a billion dollars, you know they were active.
Foreign investor flow data from the Tel Aviv Stock Exchange and the Ministry of Finance. If foreign investors are piling into shekel denominated bonds, that's the carry trade in action. If those flows suddenly reverse, it could signal that the speculative positioning is unwinding. That's the canary in the coal mine for whether the intervention is working strategically, not just tactically.
Reserves tell you what the central bank is doing. Foreign flows tell you what the market is doing. Together, they tell you who's winning.
Right now, the flows data suggests the carry trade is still attractive. Israeli interest rates are at four point five percent. US rates are expected to decline further. That differential keeps the shekel under appreciation pressure. The Bank of Israel's intervention may have bought them a few weeks, but unless the interest rate differential narrows — either through Israeli rate cuts or a shift in Fed policy — the pressure will return.
Which is exactly what we saw with Japan. The intervention works until the fundamental driver reasserts itself.
That's not a failure of intervention. It's just the reality that central bank operations in FX markets are tactical, not strategic. They can smooth the path, but they can't choose the destination. The destination is set by interest rates, growth, inflation, and geopolitics. The intervention just determines whether you get there by limousine or roller coaster.
To answer the original question directly: how much institutional money does it take to move the shekel? A single macro fund deploying two hundred to five hundred million can cause a noticeable wobble. A coordinated move by several funds totaling a billion or two can shift the rate by a percent or more. And the actors doing it are hedge funds running carry trades, momentum algorithms, asset managers rebalancing, and corporate treasuries hedging — all of whom can look like speculators depending on your vantage point.
The Bank of Israel's eight hundred and one million was a proportional response that said: we see you, we can match you, and we have two hundred billion more where that came from. Whether that's enough depends on whether the fundamentals validate their stance. If they do, the intervention will be remembered as a well timed tactical success. If they don't, it'll be remembered as an expensive delay of the inevitable.
Which is pretty much the story of every central bank intervention ever.
The details change. The shekel instead of the yen, algorithms instead of phone calls, June twenty twenty six instead of September twenty twenty two. But the fundamental dynamic — a central bank with finite reserves trying to bend a market with infinite appetite — that's as old as foreign exchange itself.
Now: Hilbert's daily fun fact.
Hilbert: The ancient Maya of Belize used fermented octopus ink as a unit of trade. One clay vessel of aged ink, roughly two liters, was worth the equivalent of twelve bushels of maize at the market in Caracol around the year four hundred. That's about one point seven bushels per modern fluid ounce.
I have so many questions, and I suspect the answers would only generate more questions.
I'm stuck on the logistics of octopus ink fermentation in the Yucatan. That's a supply chain problem I had not previously considered.
This has been My Weird Prompts. Thanks to our producer Hilbert Flumingtop. If you enjoyed this episode, leave us a review wherever you listen — it helps more people find the show. I'm Corn.
I'm Herman Poppleberry. We'll be back next week with another prompt.
