7 ms·
> They are roughly the same price in gold as they were in the 1970s So you're implying that there was no inflation between 2011 and 2022 (gold prices were basi
by qwytw 3y ago
> They are roughly the same price in gold as they were in the 1970s
So you're implying that there was no inflation between 2011 and 2022 (gold prices were basically the same) or that prices increasing 8 times or so between 2000 and 2011 because gold got a lot more expensive?
Gold is just a random commodity affected by market supply/demand just like every other commodities (or bitcoin). Implying it's some sort of a "hard currency" or can be used to compare prices of goods/services/housing over long periods of time is just absurd.
> when you realise that every year the banks are collecting interest on every dollar/euro/pound etc
That doesn't work that well when the real interest rates (i.e. after you subtract inflation) are negative or close to negative (as they were in the Eurozone between around 2012 and 2022).
- vondur 3y agoI'm guessing it's that the US Currency was backed by gold and later silver and the central banks couldn't print currency without some precious metal to back it up.
- qwytw 3y ago> was backed by gold and later silver It's the other way around. US switched from a bimetallic system to the gold standard in 1873. > I'm guessing it's that the US Currency was backed by gold Well sort of, only until 1971 (or to some extent 1933) though. So you maybe could do that if you wanted to compare the price of housing in the 1870s and 1920s, certainly not between 1970 and the 2020s.
- npoc 3y ago> Gold is just a random commodity affected by market supply/demand just like every other commodities (or bitcoin). Implying it's some sort of a "hard currency" or can be used to compare prices of goods/services/housing over long periods of time is just absurd. It's sensible to use gold or housing as a measure of value - they are both extremely mature markets with a relatively constant supply/demand ratio. It's no coincidence that after 50+ years they are still the same value relatively to each other. I'd argue that using the dollar as a measure of value would be absurd, given that it's demand is relatively constant but it's supply is doubling every decade. > That doesn't work that well when the real interest rates (i.e. after you subtract inflation) are negative or close to negative (as they were in the Eurozone between around 2012 and 2022). The banks are playing a much longer game. While the interest rates are held low, they increase the number of people indebted to them. When they turn the interest rates back up, they get their reward. I recommend reading the short book "The Great Taking" (https://thegreattaking.com/ https://thegreattaking.com/)
- qwytw 3y ago> It's sensible to use gold or housing as a measure of value - they are both extremely mature markets with a relatively constant supply/demand ratio Look at how the price of gold changed between 2000 and 2020. It's not at all constant (it's actually more volatile than the dollar) > I'd argue that using the dollar as a measure of value would be absurd, Perhaps. Still less absurd than using gold for that. > When they turn the interest rates back up, they get their reward When the interest rates go up the price of bonds goes down. If you're holding a lot of bonds and interest rates go up you're certainly not in a good position (that's how the Silicon Valley Bank went bankrupt). Just basic math. Of course if we're talking about variable rate debt then you do have a point (however almost all household debt in the US is fixed rate, it varies by country though).
- pdonis 3y ago> Look at how the price of gold changed between 2000 and 2020. It's not at all constant (it's actually more volatile than the dollar) It's pretty rich for you to be using the word "absurd" to describe what other people are posting, when you come out with this. Let me translate into actual plain English: the price of gold in dollars has gone up quite a lot between 2000 and 2020. (But note that it still hasn't reached the peak value it reached around 1980, after the huge inflation of the 1970s.) That is not because anything about gold has changed: it has the same industrial and other uses (such as jewelry) as before. The reason the price of gold in dollars has been so volatile is that dollars are volatile--because the government keeps monkeying with them. From 2000 to 2020, the government was printing dollars. In the early 1980s, by contrast, the government was destroying dollars (under Volcker), and the gold price in dollars dropped significantly. Not only that, but this observation isn't limited to gold. It's true of, well, basically everything that isn't money. For instance, houses, as npoc observed. And cars. And food. And clothes. And... The obvious conclusion from observations like this is that what is absurd is using dollars as any kind of measure of value, because the thing they are actually measuring has nothing to do with value to us, producers and consumers, and everything to do with how the government is messing with the money supply. Any commodity that has an actual use independent of government manipulation of the money supply--gold, wheat, whatever--is a better measure of value than money.
- KarlKode 3y agoNot arguing for the OPs argument but observing a pattern over a longer timespan doesn’t mean that the pattern can be observed during the whole timespan. At the same time IMHO you’re right in pointing out that gold is a poor measurement of the true cost of a house.
- paulddraper 3y ago> Implying it's some sort of a "hard currency" or can be used to compare prices of goods/services/housing over long periods of time is just absurd. It goes up and down, but it's the most stable denomination of value that exists. Yes?
- the_gastropod 3y agoAs measured by what? Volatility? Absolutely not.
- bsdpufferfish 3y agoRelatively constant demand, and difficult to increase supply.
- qwytw 3y ago> Yes? Not really? Adjusted by inflation it's very volatile and not at all stable. Gold now is worth ~7x more than in 1970 but 25% less than in 1980, yet 4x more than in 2000 but still less than in 2011... There were some periods since 1970 where the (CPI adjusted) price of gold was relatively stable but that's certainly the exception and not the rule. > but it's the most stable denomination of value that exists You might use oil instead? It's not that much more unstable than gold.. Or wheat?
- pdonis 3y ago> So you're implying that there was no inflation between 2011 and 2022 (gold prices were basically the same) He doesn't mean the price of gold in dollars, he means the price of houses in gold. In dollar terms, he's just saying that the price of gold and the price of houses have inflated by the same amount. > Gold is just a random commodity affected by market supply/demand just like every other commodities (or bitcoin). Implying it's some sort of a "hard currency" or can be used to compare prices of goods/services/housing over long periods of time is just absurd. Not as absurd as printing money and then spending it on things that nobody wants, and then claiming that that "stimulates the economy". Which is what our current monetary regime has been doing for decades now.
- qwytw 3y ago> In dollar terms, he's just saying that the price of gold and the price of houses have inflated by the same amount. They might be saying, it's just not really true: http://www.goldchartsrus.com/chartstemp/free/USHomePricesAU01.php http://www.goldchartsrus.com/chartstemp/free/USHomePricesAU0... In gold houses are about 3x cheaper now than they were back in 1970. Yet in the mid 2000s they were even more expensive than in 1970. What can we make of that besides that the price of gold is very unstable and increase in money supply is not the primary reason of that? (e.g. compare 2000s and 2010s..) There is nothing special about gold, it's just a highly speculated commodity with very volatile price and that's it. We might as well do the same experiment and use the prices of oil/wheat/etc. and it would make a lot more sense.
- azinman2 3y agoWe also mine gold every year adding to the supply, thereby reducing the price of the current supply.
- numpad0 3y agosuper cherry picking and definitions bending: > no inflation between 2011 and 2022 iPhone 4S, iPhone 14 > increasing 8 times or so between 2000 and 2011 Pentium III @ ~1GHz 1C/1T, Sandy Bridge @ ~3.7GHz 4C/4T
- qwytw 3y agoWell if we're using the price of gold to measure that there was no inflation (which of course wasn't really the case). To be fair examples are pretty bad, the iPhone 4S and iPhone 14 are not equivalent products, you can get more capable smartphones for $100-200 or less these days indicating that their prices actually went down quite a bit over the period.
- numpad0 3y agoIt's flawed, but 1x1GHz -> 4x4GHz is literally 16x upgrade, and there was also loss-leader $399 iPhone SE in 2022 much better in performance than $649 iPhone 4S. The experiences(here comes the dodgy part) are way closer between two iPhones than the gap between PIII and Sandy Bridge.
- throwaway14356 3y agolet me try something else: If you take out a lone you get money in exchange for signing a contract that forces you to pay everything back [say] 3x over 30 years. The party giving you the money loses nothing, they trade one kind of paper for the other. They get more money in return because there is a small risk you wont pay. (if they get 1/3 they have their money back) This would be sensible if the bank had this money. They don't, they bring nothing to the table and give you freshly minted fiat. This is only possible because you've signed on the dotted line. If you are buying a house the risk you wont pay is tiny! Government could easily force you to get insurance and give you an interest free lone that you won't need to pay back. This would bring some serious fiat into circulation. Ask Gaddafi what happens next.
- wakawaka28 3y agoPrice increases come in fits and starts. Gold is not entirely arbitrary. It is a commodity with a restricted and steady supply. As such it can be used as a measuring stick to judge how distorted prices are getting. Mind you, (price) inflation is measured by the government as a relative change in a cherry-picked basket of goods that changes over time. It's not honest and does not accurately account for changes in credit conditions.
- qwytw 3y ago> As such it can be used as a measuring stick to judge how distorted prices are getting It can. It would just be a pretty bad idea to use it for that since it would indicate that prices where consistently falling during the 80s and 90s amongst other. I mean the nominal price of gold in 1980 was exactly the same as in 2007... I can barely think of a worse measuring stick (we could just use the price of oil instead? It's hardly less stable). > inflation is measured by the government as a relative change in a cherry-picked basket of goods that changes over time perhaps. Seems fairly tangential, unless you're implying that governments generally tend to severely underreport inflation most of the time.
- wakawaka28 3y agoYou're right I guess. The price of gold is manipulated heavily by the banks so it is not a perfect measuring device. There is literally a Federal Reserve memo on Wikileaks describing how the government would create FUD about gold through the futures market. But the supply of it is nevertheless fairly constant compared to just about anything. People have remarked that an ounce of gold today buys about the same as it did thousands of years ago. I think they are right. The price of oil is clearly less stable than the price of gold. You could do better if you compared many goods in a basket, perhaps, but you just can't trust the government to do it right. >perhaps. Seems fairly tangential, unless you're implying that governments generally tend to severely underreport inflation most of the time. Yes that is exactly what I'm implying (or stating directly). I am mainly talking about the US government because I don't know how other governments calculate their rates, but it is probably about the same everywhere. I think in recent times you could do well by doubling their number. The US government also misrepresents unemployment. For example, if an unemployed engineer gets a job at Starbucks one day a week while applying for a new gig, he is not counted as unemployed anymore. He is also not counted if he is unemployed for any longer than 6 months. Perhaps we need to report them all as unemployed along with the median time to get a job, and/or have a separate number for skilled workers that does not exclude them if they are forced to take an inadequate job.