Death by Derivatives
damninteresting.com
damninteresting.com
> Michael Lewis details one such practice in his book Flash Boys, where market makers use superior networking technology to change the market price of a security after a customer sees it and places an order, forcing them to trade at an inferior price.
This is a potentially misleading description of HFT algorithms. HFTs cannot see a customer’s order before it executes on an exchange. They cannot “jump in front of” an atomic order. HFTs only have insight like this at the inter-exchange level, i.e. in latency arbitrage. But by default a customer’s order won’t be routed to multiple exchanges unless there is insufficient liquidity to complete the order on the first exchange (which, if that’s routine, means you’re likely an institutional customer). Alternatively, if they submit several orders, HFTs will (logically) adjust their prices in response to the new order book activity, which would result in price slippage for the customer.
Market makers can lose a lot of money under certain market conditions. Their profit margins are rather small. Market makers are practically heros.
What defines market making is the fact that you buy and sell the same commodity at the same time, taking advantage of the spread to make money.
Two other rapid transaction arbitrage strategies involve market manipulation, in which trades are placed that are designed to fool other trading algorithms into believing that the price is going to move a certain way, and statistical arbitrage, which searches for temporary price differentials across exchanges.
Neither of these two strategies create liquidity on the market, that is, offers more opportunities for anyone who wants to buy or sell to do so. Instead they seek to create, and/or exploit opportunities.
Normal market making too could be seen as exploitation. Suppose the market is crashing. Without market makers the market ceases to function and uncertainty worsens. With market makers the market functions and crashes as much as it wants to. So your stock holdings lose value and here's a market maker glad to help that happen. How is that not "exploitation"?
Politics is a war of words, I get that. I guess I should be happy that nowadays we don't punish market makers.
You need not be a formal market maker to increase liquidity. Technically speaking even retail customers will receive small transaction fees for increasing liquidity in certain circumstances where it's particularly lacking. More generally speaking, any strategy that rapidly connects two parties is market making, even if the two parties are not immediately next two each other in the strategy's trading queue.
> Two other rapid transaction arbitrage strategies involve market manipulation, in which trades are placed that are designed to fool other trading algorithms into believing that the price is going to move a certain way
Market manipulation is not a good term to use for this, and it's a fairly serious accusation to levy against a trader. Market manipulation is an illegal activity that typically involves cornering or collusion. More abstractly, this is a net improvement in price discovery, because one party notices that another party can be fooled (which means their algorithm has a glaring pricing inefficiency). In repeatedly taking money from the other party, the other party becomes incentivized to stop executing that strategy. Once there is no counterparty to take that trade anymore, the inefficiency has been traded away, and the market is now (likely imperceptibly) more efficient.
> Neither of these two strategies create liquidity on the market, that is, offers more opportunities for anyone who wants to buy or sell to do so. Instead they seek to create, and/or exploit opportunities.
This creates liquidity. Any situation that increases trading activity increases liquidity by definition. You do not need purely symmetric market making to improve overall liquidity - if many low latency trades are being executed, and they are finding counterparties, the liquidity of that market is increasing as a result of the volume. HFTs do not buy and hold, which means they are rapidly connecting buyers and sellers, even if it individual securities are bought and sold asynchronously.
I suspect too that HFTs do not buy and hold simply because the cost of holding must be very high for an HFT, e.g., transaction fees may be higher when you mean to hold, and anyways, whatever you're holding is burning a hole in your pocket, much like cash does for an investor -- it represents a missed opportunity.
I also don't see how offers/bids that no one wants to take exploit anything. If they are unreasonable then they are noise that a) can be filtered, b) costs the HFT something. If they are reasonable then they may end up in an actual trade (improving liquidity). Either way they help price discovery!
As you said, market making is super low profit.
HFT certainly used to be a fairly rich source of profit. Not sure if what it's like these days.
This is all probably just a coincidence.
Now that doesn't matter so much if you're essentially using those high volume trades as an alibi. Zero profit is fine.
All of this is an open secret. Hell, it's why dark pools exist. Those mean ungrateful dark pool customer fucks apparently decided that they didn't want all of that liquidity that HFTs were generously providing when they bought and sold. So rude.
I'm pretty sure this is what bid stuffing was all about.
I guess it's technically correct but it's not a particularly meaningful distinction given the context.
I fear the day that Bitcoin & friends come crashing down.
Might have to try it again :-)
The commonly stated notional value of the derivatives market is sort of a nonsense number. It’s based on (very roughly speaking) the number of derivatives contracts times the value of their underlyings. This means that if you had, let’s say, a billion binary contracts that paid out one cent if Apple was above $180 in one month, this would have a notional value of like $175B, even though the maximum amount of money that changes hands is $10M and the market value of the contracts is even less than that.
I looked actual estimated market value of all derivatives contracts a while back and it was something like 1.5% of the stated notional value. It’s still really big, but iirc substantially smaller than “normal” investments like stocks and commodities. Not saying the article’s wrong, it is indeed trillions of dollars even after accounting for this. Just an interesting thing that isn’t usually mentioned.
My grandpa can’t comprehend how bitcoin has “created value out of nothing” but really the finiancial world has been doing it for ages.
At first they were just a safe place in which to store your valuables (basically they were goldsmiths renting out vault space)
Then someone (I think dutch) realised that most of the time people kept there cash in one place. Which meant that instead of raising money explicitly for lending, they could "double enter" the cash, which meant any deposits to the bank were lent out immediately.
Commodity futures give other FD's a bad name.
Well, not really.
Financial instruments like stocks and bonds are real legal claims on the real assets of an entity. They are not creating value out of nothing, they are transferring ownership rights.
Derivatives are a zero-sum game. There is no value out of nothing, the money you make is your counterparty's loss.
Seignorage is what the government does when it creates dollars (or when bitcoin is created). That truly is creating money for nothing.
>Financial instruments like stocks and bonds are real legal claims on the real assets of an entity. They are not creating value out of nothing, they are transferring ownership rights.
While an instrument making a claim to a real asset can't create value from nothing, it can create value from almost nothing. For example if the going price for a particular beanie baby is $100,000 a huge amount of speculative value has been created from a very small (basically nothing) real world value. In general it's poor speculation that causes problems, not necessarily the nature of the asset itself
If I pay $1 for a beanie baby the manufacturer presumably captures the value of the difference between where they value labor and materials (< $1) and $1. I capture the value of the utility a beanie baby gives me (>=$1) less $1.
If I then mark the beanie baby on my books as worth $1 million, because that's where I have seen other similar beanie babies trade, on my balance sheet I have captured $9,999,999 because I was able to get real property the market values at $1 million for $1. That is ludicrous, but it is real (though unrealized).
Now I could sell it for the $1 million, and my gain would realized, and it would equal the difference between my cost and the (possibly irrational) utility the buyer placed on that beanie baby. That the buyer irrationally believes that a beanie baby has value beyond that of a toy, as a store of value, or a possible rising speculative asset is not prudent or wise, but it is real.
If I issue some kind of "IOU" certificate, redeemable for a rare Charizard card, and people trade those IOUs instead of redeeming them (good thing because I don't have those Charizard cards in my possession at the moment), then I'm basically expanding the supply "things that people trade and commonly use to pay for goods" (i.e. money/currency).
I didn't create any new Charizard card, but as long as people don't ask me to redeem them, it's roughly as though the market had 10 more copies of that card circulating.
There isn't more "value/wealth" created, but there is now more "money/currency" circulating. You can imagine how something similar is happening with derivatives.
- A deposits $300k
Total deposits: $300k, total lendings: $0k, total lendable: $270k
- Bank lends B $270k to buy C's house
- C deposits $270k from the sale of his house.
Total deposits: $570k, total lending: $270k, total lendable: ($570k * 0.9 = $513k, $513k-$270k = $243k)
- Bank lends D $240k to by E's house
- E deposits $240k from the sale of his house
Total deposits: $810k, total lending: $510k, total lendable: ($810k * 0.9 = $729k, $729k - $510k = $219k)
- Bank lends F $210k to by G's house
- G deposits $210k from the sale of his house
Total deposits: $1,020k, total lending: $720k, total lendable: ($1020k*0.9 = $918k, $918k - $720k = $198k)
In this process, the bank has turned $300k into $1,020k, and has loaned out $720k, while never loaning out more than they have deposited. That's the magic of fractional reserve banking.
So I don't think it makes any sense to say they turned $300K into $1,020K. C, E, and G already had existing houses, and the series of events isn't going to change the amount of deposits or loans in the world as a whole. It doesn't give the bank a profit of $720K. They get interest on $720K, but they have to pay interest on $1020K so obviously they need a spread in rates to do it.
As far as I can tell, you're just describing the bank getting a larger balance sheet and they've got to raise more capital to do it which limits profits.
The total money supply has grown though. From an initial $300k, they now have 4 people who all, looking at their bank statements, have deposits totaling $1M. If they all tried to withdraw that at the same time, there'd be trouble, but fractional reserve banking rests on the assumption that they won't.
And yeah, there's a pretty significant spread on interest rates between deposits and loans.
The GP post was talking about "creating a dollar". This is precisely how banks "create dollars". As those loans are repaid and the deposits are withdrawn, they're subsequently "destroying" those dollars from a money supply perspective.
And yes, the banks are controlling the money supply. See https://en.wikipedia.org/wiki/Fractional-reserve_banking for more detail.
Edit: why do you say withdrawing money destroys it? It removes it from a bank's balance sheet, but not from the economy...
The bank's assets and liabilities balance out. The new deposit is a liability and the loan is an asset. The bank's total deposits rise by $1000. Because normal deposits count towards M2 money, and this new deposit appeared out of thin air, the M2 rises by $1000. New money has been created.
Banks do have to make sure that they meet their capital and reserve requirements, e.g. if they find themselves short they can increase their reserve by taking out a loan from the central bank.
Institutions such as banks and credit unions are also able to create money 'from nothing', via lending.
When they make a loan, it's not as though they deduct the amount loaned (or any part of it) from someone else's account, and typically they're allowed to make loans totaling many times (9x? 11x? I forget) the amount of deposits they possess.
I think I must have been assuming a situation along the lines of "bank loans out 90% of its deposits, and then all those debts are moved to other institutions", in which case I expect the bank would be left with 10% of its original deposits plus the loan agreements, the total value of the latter being equal to ~9x the value of the remaining deposits.
> Because banks hold reserves in amounts that are less than the amounts of their deposit liabilities, and because the deposit liabilities are considered money in their own right, fractional-reserve banking permits the money supply to grow beyond the amount of the underlying base money originally created by the central bank.
https://en.wikipedia.org/wiki/Fractional-reserve_banking
It is literally just paperwork.
Fractional-reserve banking is the greatest scam ever, or a pillar of our economic system, depending on whom you ask.
In any event...
> hyperinflation is often associated with some stress to the government budget, such as wars or their aftermath, sociopolitical upheavals, a collapse in export prices, or other crises that make it difficult for the government to collect tax revenue.
~ https://en.wikipedia.org/wiki/Hyperinflation
Banks generally give loans to people who they believe are very likely to repay them. Typically this means that you get a loan if you have a good plan for repaying it and a history of doing so. (Leading to the old complaint, "The only people who can get loans don't need them.")
These are two different kinds of "money creation".
IANAEconomist
Reserves are not the amount of assets the bank has on its balance sheet - if they had less assets than their liabilities (aka deposits), they would be insolvent.
Capital ratios and reserve ratios are different concepts - just as solvency and liquidity are different.
I am open to the idea that fractional-reserve banking is a scam, because of my experience of human nature, especially when money is involved.
Certainly the history described in the wikipedia article makes it sound like a clever trick the old goldsmiths came up with, and no one the wiser until time and custom had rendered it respectable.
> the goldsmiths observed that people would not usually redeem all their notes at the same time, and they saw the opportunity to invest their coin reserves in interest-bearing loans and bills.
The trick works fine until it doesn't.
> If creditors (note holders of gold originally deposited) lost faith in the ability of a bank to pay their notes, however, many would try to redeem their notes at the same time. If, in response, a bank could not raise enough funds by calling in loans or selling bills, the bank would either go into insolvency or default on its notes.
This activity is different from mining, or farming, or manufacturing, or shipping, or goldsmithing. It's a kind of consensus hallucination (or a hoodwink) that worked and kept working.
Now we could argue all day whether the bankers né goldsmiths deserve to make money this way, but I'm not interested. If it really bothered me, I would start a bank. ;-)
I'm not financially sophisticated. To me, if there's more cash than gold, something's fishy. :-)
-------
edit: (Apparently I have nothing better to do this morning.)
Let me see if I can clarify my position:
Long ago my friends got into playing Magic the Gathering. They tried to get me to play and I said I would but I was going to make my own cards. (Make photocopies of the real cards and laminate them and play with those.) To me it seemed totally obvious and reasonable. After all, in D&D no one tries to make you only use the official dungeon modules, eh? But my friends wouldn't hear of it. I got kind of mad at them and told them, "You know there are no real wizards at Wizards of the Coast, right?" But they were baka as the Japanese say. Fools.
For what it's worth, this is my attitude to banking: Neat trick, I can't believe people go for it, but whatever.
As long as anyone can open a bank and get in on the "scam" I can't really condem it. Consenting adults and all that.
If a bank has too many people demanding their deposits back at once, that's a liquidity crisis. It's not insolvent. A bank can be perfectly solvent, in that it has more assets than liabilities, and still not have enough reserves to satisfy people right now.
a) Imagine a coin tossing game. We both put in a dollar and flip a coin; heads you win both dollars, tails I win both dollars. Truly zero sum.
b) Consider a vendor selling me a $5 hamburger for $5. They lose $5 worth of goods and gain $5 worth of cash; I lose $5 worth of cash and gain $5 worth of goods. In an accounting and physics sense this is zero sum; no dollars or molecules were created or destroyed, the universe just slightly rearranged itself. But in economic terms, this isn't a zero-sum transaction; the purchase created value since at that exact moment I valued the hamburger more than the cash (due to hunger), and the vendor valued the cash more than the hamburger. We both improved our situation and created value "for nothing". After eating the hamburger, I satiate my hunger, and now value a second hamburger less than $5, which prevents a second trade from happening for a few hours.
c) Let's say I'm a wheat-producing farmer and you're a wheat-consuming baker, and there are thousands of others like us in the economy so neither of us really has any effect on wheat prices. When the wheat price is up, I benefit, and when the wheat price is down, you benefit. At this point we have a zero sum situation. Now let's say we're both reasonably content with the wheat price being what it is right now, so we sign a "futures" agreement for me to deliver wheat to your bakery for the next year, locked in at the current price. By signing the contract, I'm choosing to forgo extraordinary profits next year if the wheat price happens to go up, but also insuring myself against extraordinary losses if the wheat price happens to go down. Your end of the bargain is the same thing in reverse.
Like the hamburger example, in an accounting sense this is zero-sum. My potential profits are exactly equal to your potential losses, and vice versa. But in an economic sense, there was another quantity that was affected in this transaction: risk. Both of us reduced our wheat price risk to zero for the next year. Risk is something people intrinsically dislike, so it has disutility (i.e. negative value). Eliminating a negative therefore adds value.
As a concrete example, let's consider what would happen if you couldn't sign the futures contract above--let's assume no modern banking or finance at all for this example--and had to just accept wheat price fluctuations. Overall, on average, the highs and lows should more or less cancel out. But on a short-term basis it's not unlikely at all to have a run of 3-5 years of bad prices (too low for me, or too high for you). To survive this potentiality, we'd both have to set aside a considerable rainy day fund just in case I had to cover 3-5 straight years of depressed profits or outright losses. Most of the time this cash will be unused, but the mere threat of a prolonged unlucky spell means we can't use that cash to buy more land or tools. The net effect on all farmers and all bakers is that everyone has a large chunk of their wealth sitting around in a completely unproductive state, which ultimately means less loaves of bread are produced and society is relatively impoverished.
Likewise, other financial derivatives like options, swaps, etc., also create economic value. They might be zero sum in a cash sense, but they affect some more-abstract quantity like "risk" or "leverage" and therefore create value for both buyer and seller.
My original point was that the finance whizzes aren't creating money.
Your gain/loss from a derivative contract is paid by/to your counterparty.
As opposed to a stock, which is a legal claim on the assets of an enterprise. When an IPO happens there is no magic in which wealth is being created -- rights to the enterprise are being sold.
(This is distinction with a typical ICO, in which what is being sold is often unclear or simply worthless. That is seigniorage -- or simply a scam.)
If you've written an IOU, you've created a kind of currency. Same for arcades, carnivals, etc. Any time you acquire a service or good in exchange for a promise of a service or good in the future (and that promise could also be in a tangible thing like gold, seashells, or bottles of Tide, but often it's just word of mouth), you've created a kind of currency. The real ticket comes when you can convince third parties to accept your promise and trade goods in exchange for it.
But alternative currencies have a really long history, as old as society and maybe older. They tend to arise during depressions when fiat currency is hard to come by. I don't think it's by accident that Bitcoin was first published in 2008 and released in 2009, during the Great Recession. This is liquidity being popped out of the ether due to demand for more currency (in spite of pumping by the Fed). And this is why I never really bought the "infinite" deflationary story for Bitcoin: people can just start another cryptocurrency if there's actual demand for it. Of course, that doesn't take into account irrational speculation...
The current ramp-up in cryptocurrencies is occurring now, during a relative economic boom. This is pro-cyclical and probably bad. It'll eventually pop, but no one knows when.
It's pretty obvious that free markets (with a few guardrails) work better than anything else we've tried so far. Yet there's so much overhead spent on managing the system, one wonders how anything ever gets done...
On top of that, the currency that has been given legal tender status by government, and the currency used as the basis for calculating capital gains tax, just happens to be issued by the central bank that’s willing to monetize the government’s debt. What a peculiar coincidence.
This doesn't make sense. If every dollar of their revenue created more than a dollars worth of value for the rest of the economy, then it's a net win. To put it in a different way, consider the economic losses if you couldn't get cheap mortgages, loans, etc. from financial institutions.