I've been thinking about this since I read your earlier piece on this and I think I found the error. There's the efficient market hypothesis (EMH )which is that a purchase or sale in a market only moves the price by the bid-ask spread multiplied by the size of the order. For deep markets and small orders the effect of the inflow or outflow is zero. But in reality there's a multiplier on that; for stocks it's about 3 to 5. Also the bid ask spread can be considerable in illiquid markets. So overall the effect of an inflow or outflow is real. What do you think?
When choosing between a hypothesis and reality, I'm going to take reality all day. So the fact that you need to add some nebulous, poorly defined "multiplier" to your hypothesis to fill the gap between hypothesis and reality just means that your hypothesis is incorrect.
Here's your error. As long as there is a bid-ask spread, there are no transactions and therefore no market price. In order for a market price to be established, a buyer or a seller must move their bid or ask to a level that will be accepted by the other party. It's the act of moving the bid or ask that creates a sale and therefore changes the market price, not the size of the sale.
The bid-ask spread can be $0.10, and if sellers decide to raise the ask $1,000, a new market price will be established once buyers move the bid $1,000.10. That will happen with only a single transaction of "inflow", regardless of the size, which will make the "multiple" on that "inflow" some ridiculously large number.
The fact that this extreme example doesn't happen often doesn't invalidate the principle, I'm just using an extreme example to illustrate that your hypothesis fails to accurately describe reality in every case, which is why it requires a "multiple" rescuing device, while my explanation describes reality even in extreme outlier cases.
The multiple was measured experimentally. It's a weight on the equation.. Any model has weights on the terms of the equation; I would not consider it a crutch. A equation for a chemical reaction can be the same except for different weights depending on the specific elements involved.
I think you're misunderstanding the market price. It falls between the bid and the ask price. It isn't correct to say that no transactions occur if there's such a spread, it simply means the market is illiquid. You can look at the bid ask spread of a small cap stock vs. a large index ETF. The former will have a large one and the latter will have maybe ten cents of difference. Both have regular transactions.
So you're saying short term holders were buying at a rate one or two sigma higher than the mean, and the price still fell? Shouldn't it rise with all those inflows from short term holders?
I see how you inverted, but no it doesn't follow the way you say. The LTH sellers are willing to sell at the market price (e.g. 100k) or between the asking price and the market price (100.2k and 100k). As LTH list their BTC sell orders, they are reducing the bid price and consequently the market price.
Yes, and of course sentiment isn’t a point, it’s a curve. Two curves, in fact (supply and demand). “Inflows” increase when there is a lot of overlap in the curves, which tends to happen when there is more diversity in views about the value. For example, if something changes people might update their views and the curves adjust accordingly.
It’s important that you point out that if we are talking about inflows of BTC into an ETF (as opposed to inflows of USD), the supply is non-ETF holders and the demand is the ETF holders.
In futures markets there are locals and paper. Locals buy and sell from each other all day and the price reverberates but doesn't shift significantly in one direction or another (because they usually don't have the liquidity to hold positions overnight).
However, when paper comes in (the money flow) it will move the price. Paper (insititutions) don't have to close their position later that day. They might not close it for months. Money flow is external to that market, and thus moves it. Quite substantially in fact.
All markets work approximately the same as this; external money to each market moves it. That's why money flows are very important and large ETF providers monitor this very carefully to design their next ETF.
In your analogy about sellers just not selling for less than 200k, that's functionally the same as an aggressive buyer taking every sell order at once up to 200k. The price would shoot to 200k until new sell orders are made.
It's only functionally the same in the sense of the effect on price. It's actually completely different in the sense that no transactions occur versus your scenario in which lots of transactions occur. That's exactly the point I'm making: if price can move the same with or without transactions occurring, then transactions occurring are not the cause of price movement. That's the crux of my argument, and the most distilled explanation of why "inflows" are economic hocus-pocus.
So let's go through the steps to make sure I understand. First the sellers remove their orders up to 200k all at once, let's say it is because there's an announcement of a sovereign buying BTC for that price. The ask explodes upwards. The marker is either frozen due to a massive spread resulting in no transactions and therefore no market price, or the buyers also adjust their bids upwards. Both sides would eventually adjust until they met in the middle somewhere because there is always a desperate buyer and a desperate seller. There would be transactions at the price and the inflow that day is measured in the net difference in price from the previous day (100k) times the number of transactions (should be about the same as the previous day). So the inflows would show as large because sellers are getting a much larger sum than they paid for the BTC.
Inflows are calculated as price times volume. Whether it's an inflow or an outflow is determined by whether the price is higher or lower than the previous day. So if the daily volume is a billion dollars, and the price is up ten dollars today, that's a billion dollars of inflow. If volume tomorrow is a billion dollars and price is down a thousand dollars, that's a billion dollars of outflow.
If your scenario happens and offer price rises a hundred thousand, and then one single transaction for a hundred dollars occurs at the higher offer price, that's a hundred dollars of inflows.
Whether more transactions occur that day or not, and whether total volume ends up being a billion dollars for the day or stays at a hundred dollars, the same point stands. The price move occurs BEFORE the "inflow" happens. That's why you can have a billion dollars of "inflows" on a day price moves ten dollars, and a hundred dollars of "inflows" on a day price moves a hundred thousand dollars. That's the point I'm trying to make, "inflows" don't CAUSE price movement. They may occur coincidentally with price movement, or they may not.
The realized market cap metric is great for looking at BTC because it is weighed by transaction size rather being the most recent transaction price times the number of BTC outstanding. Only possible because of the open ledger.
I see them as happening simultaneously as the price information enters the system by the buyer and seller transacting in the order book. Yes, the flow can be calculated using the price but one can also calculate the expected market price after a transaction. The concept of an inflow is useful because it can be a measure of how much money is needed to chew through an order book to move the price.
The price of BTC fell because of outflows. ETFs are only part of the picture. We have the on chain data to see that long term holders were selling at a rate one to two sigma higher than the mean. Summed together, we get a better picture.
So what does inflow really mean? nothing at all?
Yes, nothing at all. The dollar amount of buying and selling is always by definition exactly equal.
I've been thinking about this since I read your earlier piece on this and I think I found the error. There's the efficient market hypothesis (EMH )which is that a purchase or sale in a market only moves the price by the bid-ask spread multiplied by the size of the order. For deep markets and small orders the effect of the inflow or outflow is zero. But in reality there's a multiplier on that; for stocks it's about 3 to 5. Also the bid ask spread can be considerable in illiquid markets. So overall the effect of an inflow or outflow is real. What do you think?
So what is the multiplier? Is it 3? 5? 4.39?
When choosing between a hypothesis and reality, I'm going to take reality all day. So the fact that you need to add some nebulous, poorly defined "multiplier" to your hypothesis to fill the gap between hypothesis and reality just means that your hypothesis is incorrect.
Here's your error. As long as there is a bid-ask spread, there are no transactions and therefore no market price. In order for a market price to be established, a buyer or a seller must move their bid or ask to a level that will be accepted by the other party. It's the act of moving the bid or ask that creates a sale and therefore changes the market price, not the size of the sale.
The bid-ask spread can be $0.10, and if sellers decide to raise the ask $1,000, a new market price will be established once buyers move the bid $1,000.10. That will happen with only a single transaction of "inflow", regardless of the size, which will make the "multiple" on that "inflow" some ridiculously large number.
The fact that this extreme example doesn't happen often doesn't invalidate the principle, I'm just using an extreme example to illustrate that your hypothesis fails to accurately describe reality in every case, which is why it requires a "multiple" rescuing device, while my explanation describes reality even in extreme outlier cases.
The multiple was measured experimentally. It's a weight on the equation.. Any model has weights on the terms of the equation; I would not consider it a crutch. A equation for a chemical reaction can be the same except for different weights depending on the specific elements involved.
I think you're misunderstanding the market price. It falls between the bid and the ask price. It isn't correct to say that no transactions occur if there's such a spread, it simply means the market is illiquid. You can look at the bid ask spread of a small cap stock vs. a large index ETF. The former will have a large one and the latter will have maybe ten cents of difference. Both have regular transactions.
So you're saying short term holders were buying at a rate one or two sigma higher than the mean, and the price still fell? Shouldn't it rise with all those inflows from short term holders?
You should review my comment above about the EMH. I think it was a good explanation.
I see how you inverted, but no it doesn't follow the way you say. The LTH sellers are willing to sell at the market price (e.g. 100k) or between the asking price and the market price (100.2k and 100k). As LTH list their BTC sell orders, they are reducing the bid price and consequently the market price.
Correction: I meant to write asking, not bid price.
Yes, and of course sentiment isn’t a point, it’s a curve. Two curves, in fact (supply and demand). “Inflows” increase when there is a lot of overlap in the curves, which tends to happen when there is more diversity in views about the value. For example, if something changes people might update their views and the curves adjust accordingly.
It’s important that you point out that if we are talking about inflows of BTC into an ETF (as opposed to inflows of USD), the supply is non-ETF holders and the demand is the ETF holders.
In futures markets there are locals and paper. Locals buy and sell from each other all day and the price reverberates but doesn't shift significantly in one direction or another (because they usually don't have the liquidity to hold positions overnight).
However, when paper comes in (the money flow) it will move the price. Paper (insititutions) don't have to close their position later that day. They might not close it for months. Money flow is external to that market, and thus moves it. Quite substantially in fact.
All markets work approximately the same as this; external money to each market moves it. That's why money flows are very important and large ETF providers monitor this very carefully to design their next ETF.
In your analogy about sellers just not selling for less than 200k, that's functionally the same as an aggressive buyer taking every sell order at once up to 200k. The price would shoot to 200k until new sell orders are made.
It's only functionally the same in the sense of the effect on price. It's actually completely different in the sense that no transactions occur versus your scenario in which lots of transactions occur. That's exactly the point I'm making: if price can move the same with or without transactions occurring, then transactions occurring are not the cause of price movement. That's the crux of my argument, and the most distilled explanation of why "inflows" are economic hocus-pocus.
So let's go through the steps to make sure I understand. First the sellers remove their orders up to 200k all at once, let's say it is because there's an announcement of a sovereign buying BTC for that price. The ask explodes upwards. The marker is either frozen due to a massive spread resulting in no transactions and therefore no market price, or the buyers also adjust their bids upwards. Both sides would eventually adjust until they met in the middle somewhere because there is always a desperate buyer and a desperate seller. There would be transactions at the price and the inflow that day is measured in the net difference in price from the previous day (100k) times the number of transactions (should be about the same as the previous day). So the inflows would show as large because sellers are getting a much larger sum than they paid for the BTC.
Inflows are calculated as price times volume. Whether it's an inflow or an outflow is determined by whether the price is higher or lower than the previous day. So if the daily volume is a billion dollars, and the price is up ten dollars today, that's a billion dollars of inflow. If volume tomorrow is a billion dollars and price is down a thousand dollars, that's a billion dollars of outflow.
If your scenario happens and offer price rises a hundred thousand, and then one single transaction for a hundred dollars occurs at the higher offer price, that's a hundred dollars of inflows.
Whether more transactions occur that day or not, and whether total volume ends up being a billion dollars for the day or stays at a hundred dollars, the same point stands. The price move occurs BEFORE the "inflow" happens. That's why you can have a billion dollars of "inflows" on a day price moves ten dollars, and a hundred dollars of "inflows" on a day price moves a hundred thousand dollars. That's the point I'm trying to make, "inflows" don't CAUSE price movement. They may occur coincidentally with price movement, or they may not.
The realized market cap metric is great for looking at BTC because it is weighed by transaction size rather being the most recent transaction price times the number of BTC outstanding. Only possible because of the open ledger.
I see them as happening simultaneously as the price information enters the system by the buyer and seller transacting in the order book. Yes, the flow can be calculated using the price but one can also calculate the expected market price after a transaction. The concept of an inflow is useful because it can be a measure of how much money is needed to chew through an order book to move the price.
The price of BTC fell because of outflows. ETFs are only part of the picture. We have the on chain data to see that long term holders were selling at a rate one to two sigma higher than the mean. Summed together, we get a better picture.