Hacker News Reader

Comment by kragen | Hacker News Reader

Just to clarify, the dollar was gold-backed from 01944 until the end of Bretton Woods in 01971; it wasn't a fiat currency.

Also I think saying that inflation was "lower on average" is really a dramatic understatement. From independence up to the advent of the "mixed system" in 01933 (really a switch to fiat currency until 01944) inflation was probably zero on average. The same was observed over the 1300 years of the commodity-backed pound sterling (usually gold, sometimes silver or bimetallic).

I think there are plausible arguments that fiat currencies are less prone to such crises, and even that they stimulate economic growth, but I don't think it's an open-and-shut case.

View on HN

Average inflation rates over multiple lifetimes largely irrelevant. 24+% inflation for 2 years in 1863 was a huge issue. Short term deflation is often much worse than inflation. Averaging both to zero is wildly misleading.

The vast devaluation of silver after the discovery of the new world was a massive issue in Europe even if it averages out to nothing over time.

These kind of fluctuations that average out make the money useful on an immediate or long-term timeline, with the cost of medium-term fluctuations (think hold under a few months or over 5 years). Fiat as currently found in places like the US makes the money useful in the immediate or medium term timeline (think under 5 years).

I think which you prefer depends on your station in life and your spending choices. Personally when fiat hits my bank account I want to shit-can it as soon as possible, eithre for equities or commodities. If I were paid in the kind of unstable but averaging out dollar of the 1800s I would very much prefer that, rather than losing money to middle men converting to some commodity like that (such as a commodities index, silver, whatever) and then converting back whenever I need to spend.

Being able to store value as currency for long periods is a bug not a feature as it does bad things to the economy.

Lets’s say you want to store gold in your basement for bad times like a famine. Great you and everyone else has 1+ oz of gold when a famine hits, but society didn’t actually get any more food so everyone having gold is not just pointless but wasteful both to mine the stuff and to keep it protected from thieves.

Alternatively, society can try and store value as productive assets like stocks. On a day to day basis it looks similar but rather than building ever larger vaults and digging ever deeper it’s a quest to create even the smallest income stream. You still hit diminishing returns but it’s stuck with diminishing positive returns.

PS: > losing money to middle men converting to some commodity

You end up losing even more money when the government is trying to get and store enough gold to keep a gold backed currency, it’s just less obvious. Great Britten sent many ships to Canada during WWII to keep it’s gold safe. They had much much useful things to do with those ships and men.

This is a specious (ha!) argument because the total amount of effort required to safeguard precious metals is so small, because the metals themselves are so small. Gold currently costs US$55.92 per gram. Gross world product is US$80 trillion per year over 7.9 billion people; this works out to US$10.1k per person per year, or US$770k per person per 76-year lifetime. The lifetime earnings of an average person are thus 14 kg of gold, less than a liter.

It does not require a significant fraction of your lifetime to dig enough of a vault to protect a liter of gold from thieves, nor to mine it. The impact on overall economic productivity is probably similar to that of spelling "through" as "through" rather than "thru".

You can easily wear a year's wages as jewelry, or a lifetime's wages with more difficulty, which has historically often been useful when you're fleeing famine-stricken areas. Stocks have usually been less useful; protecting a stock portfolio from disaster requires active investment management and is generally infeasible during wars and famines.

It is true that collectively we need to store food, not gold, to prepare for famines. It is a mistake to leap from this fact to the conclusion that everyday people should not be allowed to save money. You might or might not be aware that until roughly the transition to fiat currency in 01971 saving money was a canonical example of responsible adult behavior in the US, much like doing your tax returns, separating your recycling, or investing in an index fund today.

I've presented a better argument for why being able to store value as currency is harmful in https://news.ycombinator.com/item?id=32258546. I'm not completely convinced by it but it's at least not completely specious.

First you can most of the benefits of gold as a currency by just buying gold and then selling it in 40 years. At a personal level a gold backed currency doesn’t change much. Anyway, if gold was actually the words medium of exchange it’s value would be vastly higher.

A reasonable estimate is the value of the worlds gold would equal the value of the words current money. As it’s a nice round number the US’s 20 Trillion M1 money supply and ~20% share of the world GDP suggests something like 100 Trillion for a reasonable estimate for the words supply of gold in that hypothetical situation.

The actual supply of gold is around 400,000,000 lb. So in that hypothetical world 1 lb of gold would in theory be worth ~100,000,000,000,000 / 400,000,000 = 250,000$/lb. Let’s again simplify to ~10x as much.

But wait if it’s worth 10x as much two things happen first it’s used less often in electronics etc which directly hurts the economy, but next it’s vastly more profitable to mine and thus the world spends more effort mining stuff that just sit’s in a vault. Except, governments can’t safely issue currency without being able to back it up with gold so they need to buy up most of this mined gold and then securely transport, store, inspect it etc.

As to keeping it in your own vault, you can secure it all you want that just makes home invasion vastly more profitable. As long as you have access to your gold at home, so does anyone with a gun to your head.

I suppose that, if you're right that a hypothetical return to the gold standard would greatly increase the value of gold, that would greatly diminish the amount of gold you need to safeguard.

There's nothing in the use of commodity money rather than fiat money that prevents fractional-reserve banking, though you seem to be suggesting that there is, since you're doing the computation based on M1 (though I think you may be using the number for M2). The St. Louis Fed estimates the velocity of M1 and M2 at 1.122, which I guess is transactions per year: https://fred.stlouisfed.org/series/M2V so I think you're probably right that gold would be about 5 or 10 times as expensive without fractional-reserve banking.

So instead of 700 grams of gold your lifetime earnings would be 70 grams.

If 5% instead of 0.5% of the cost of electronics was gold, I don't think you'd be able to tell the difference. Maybe that boosts the economic impact of this return to the gold standard to the same as spelling both "through" as "thru" and "thorough" as "thoro".

Your implied prohibition on fractional-reserve banking would, by contrast, have an enormous economic impact.

The most destructive bit was the pointless push to mine more gold. The physical volume of gold has very little to do with how expensive it is to keep safe.

Anyway, I thought M1 excluded fractional reserve money which is why it ignores money in bank vaults. Aka as you pay down a loan the M1 money supply is unchanged. Do you have a source for that?

M1 is listed at 20T here: https://fred.stlouisfed.org/series/M1SL

PS: Before May 2020, M1 consists of (1) currency outside the U.S. Treasury, Federal Reserve Banks, and the vaults of depository institutions; (2) demand deposits at commercial banks (excluding those amounts held by depository institutions, the U.S. government, and foreign banks and official institutions) less cash items in the process of collection and Federal Reserve float; and (3) other checkable deposits (OCDs), consisting of negotiable order of withdrawal, or NOW, and automatic transfer service, or ATS, accounts at depository institutions, share draft accounts at credit unions, and demand deposits at thrift institutions.

Beginning May 2020, M1 consists of (1) currency outside the U.S. Treasury, Federal Reserve Banks, and the vaults of depository institutions; (2) demand deposits at commercial banks (excluding those amounts held by depository institutions, the U.S. government, and foreign banks and official institutions) less cash items in the process of collection and Federal Reserve float; and (3) other liquid deposits, consisting of OCDs and savings deposits (including money market deposit accounts). Seasonally adjusted M1 is constructed by summing currency, demand deposits, and OCDs (before May 2020) or other liquid deposits (beginning May 2020), each seasonally adjusted separately.

>The physical volume of gold has very little to do with how expensive it is to keep safe.

I'm having my doubts about that. If I can fit all my life's savings in my pocket, I would very much prefer to do that with savings in one pocket and gat in the other (yes I know gold doesn't fit that for everyone, but lets imagine the physical volume is smaller since you said it doesn't matter). Fight to the death against violent intrusion upon your person. That doesn't incur me any material cost. If I die I won't be needing it anyway. would 100000% prefer that to sitting in the bank where some bonehead at a three letter agency can freeze or seize it at will, maybe the help of a rubber stamping judge. Or worse yet some unelected governors at the fed just choose to inflate it away.

I guess there is some argument to be made it's cheaper to keep something safe if it is either very small or very large. Small enough and you can more effectively hide it or keep it on your person. Large enough and someone will need a semi, a few men, and a big forklift or crane to steal it. Houses come to mind as something worth a lot but unpractical to steal in the conventional sense of just taking off with it.

I can't imagine the amount of costs lost in places like Argentina by people trying to find a way to keep their fiat held in a safe way.

You may be interested in the 02006 Ramallo bank robbery, in which an acquaintance of mine lost her life savings: https://www.gq.com/story/the-great-buenos-aires-bank-heist

Houses are worth a lot, but they are not worth a lot per kilogram. Consider a house consisting of four 4m×4m rooms and a floor, with 200-mm-thick floor and 3-m-tall walls made of 2.4g/cc concrete, plus a cheap sheet-metal roof. This house might cost US$100k, but its 144 m² of walls and 64 m² of floor contain 41.6 m³ of concrete. (We can disregard the weight of the roof as insignificant.) So it weighs 99.8 tonnes and only costs US$1.002/kg, 4.7 orders of magnitude less than gold. Different construction techniques can modify this somewhat but by less than an order of magnitude.

I think the lesson you should draw here is not "a large thing such as a house is impractical to steal in the conventional sense of just taking off with it" but rather "a thing worth only US$1/kg is impractical to steal in the conventional sense of just taking off with it, unless it floats". It's also unusual for people to steal pieces of a junk car chassis, for example, even though all the equipment you need is an angle grinder or a portable oxy rig.

If a thing is impractical for a thief to carry away, it is also impractical for the owner to carry it away themself, or to conceal their possession of it from thieves.

>You may be interested in the 02006 Ramallo bank robbery, in which an acquaintance of mine lost her life savings: https://www.gq.com/story/the-great-buenos-aires-bank-heist

Great read, thank you.

You're right about the amount of M1; it's US$20T in the US. Thank you.

In fractional reserve banking demand deposits are lent out. So if you print $1000 cash (whether fiat or commodity-backed), you deposit it in a bank, the bank lends out $800, someone deposits that in their bank, that bank lends out $640, and someone else deposits that in their bank, then M1 is not $1000 but $2440. If that last person withdraws the $640 from their bank and repays the loan then M1 contracts to $1800. (And that's still true if they held the $640 as currency in their mattress rather than putting it in a bank.)

So, paying down a loan does not leave M1 unchanged. You're thinking of MB or maybe M0.

Gold mining is, due to its scale, not especially destructive; the amount of gold in circulation is about 60–100 years' worth of mining at current rates. There are many environmental and human-rights problems associated with gold mines, but their total magnitude is small just because there's not that much gold out there to mine.

The physical volume of gold has a great deal to do with how expensive it is to keep safe. So does its mass. In your earlier comment, you were talking about the effort required to dig vaults, for example; this effort scales directly with the volume of the gold.

I was going to write a rebuttal, but instead I focus on an amusing byproduct of your points and points above:

1) low inflation good because long-period storage bad

2) Government mishandling of fiat policies also bad

3) 3rd party custody is expensive

The solution: radioactive commodity money of high enough density for self-custody. Maybe one with a half-life of like 20 years. Tritium is a close candidate. Keep your life's wealth in tritium in one pocket and your gat in the other.

Both short-term and long-term inflation are relevant for different purposes.

It matters a lot whether the expected average inflation rate in the future is ±½%, 4%, or 25% when you get a 30-year mortgage on a house, sign a 99-year lease, save money for retirement 45 years in the future, buy a life insurance policy that probably won't pay off for 50 years, or rent an apartment to someone under rent-control laws that can hold the rent fixed until they die (perhaps 60 years from now).

±½% compounded over 30, 45, 50, 60, or 99 years is -14%–+16%, -21%–+25%, -23%–+29%, -16%–+35%, or -39–+64%.

4% compounded over 30, 45, 50, 60, or 99 years is -71%, -84%, -87%, -91%, and -98%. So in an environment where inflation averages 4% per year, like the current US environment, such contracts are unviable even if the inflation is fairly predictable. Rent control becomes politically impossible because renting an apartment amounts to selling it. Saving for retirement requires gambling on the stock market and subsidizing the Fortune 500 with your savings.

25% annual inflation is the average we've experienced during the time I've lived in Argentina, most of my adult life. At 25% annual inflation even renting an apartment to live in becomes a major problem. The US had 24+% inflation with commodity money for two years during the Civil War; Argentina has been averaging 25% for decades now, thanks to fiat money.

The vast devaluation of silver was a quite manageable problem by comparison, except perhaps for Spain.

But it's true that volatility is a real problem for a currency.

The bigger problem often ascribed to commodity money in a growing economy is the incentive to hoard: if the economy grows 3% per year, but gold or silver (or Bitcoin) is mined at a much smaller rate, maybe the price of gold will rise in tandem with the economy. And then you can earn a 3% return just by buying gold jewelry, so maybe you won't be tempted to take the risk of investing in productive ventures (or lending to them) by buying stock and hoping for a 5% return.

My review of the Warren–Pearson index data in https://news.ycombinator.com/item?id=32258172 makes me wonder if this problem in fact exists, because it seems like gold did not appreciate significantly over the period 01776–01932, compared to their changing basket of goods. But maybe it existed in other places (maybe the US is atypical) or maybe its apparent nonexistence is an artifact of Warren and Pearson's methodology?

Reply on news.ycombinator.com