The natural state of currency is reverse devaluation, and not devaluation.
A fixed quantity of currency matched against the growing set of all goods and services naturally results in a decreasing, not increasing, ratio of currency to goods and services. This implies that a fixed quantity of currency cannot be stable in value, but holds an ever-changing value: Per unit of currency, in a fixed quantity of currency discussion case, currency's value must always grow. Per the total quantity of currency, as ratio to the quantity of goods and services, such currency's value must always decrease.
To maintain, then, both the per-unit, and total value of the coinage against the sum of all goods and services as constant values, the quantity of the currency, or the "money supply" must rise one-for-one with the growth of the material economy.
Not to get into the endogenous money theory, in which money supply can come entirely from growth in loans--versus MMT's, especially Stephanie Kelton's, emphasis on fiscal spending as source of money supply. Proponents of MMT subscribe to the endogenous money theory somehow without seeing any contradiction, by the way.
Mainstream economics doesn't seem to pin-down exactly how money supply comes out of the private economy, yet has the state dependent on said private economy supply for expenditure.
What is more clear is that mainstream economics subscribes the *value* creation as coming solely from the private economy. Now that is a bit of nonsense because obviously any efficiency discrepancies between the supposedly more productivity enhanced private sector and the public sector, in say, building a for-profit-turnpike versus a simple public highway, for example--it's not a factor of some number, but rather a fractional difference--and it's more gut-instinct than established fact that the private sector is de facto more efficient any way. Yet, none of that is necessary, as it comes down to that to say that the public sector provides *any* value at all in performing some function, whether road construction, police, military defense--to say that is to say that value creation doesn't de facto come solely from the private economy. One might say, it would be better if value came only out of the private economy, but one cannot logically state that a central government over the common wealth, *can't* create value.
To go back to my topic. The minted coin: on top of the base value inserted into it's core by the artificial means of taxation providing a base-demand for the currency (to pay the private sector's taxes with), each coin's value is little-by-little increasing, per coin, in value, as it circulates, as debt is accrued that is denominated in that currency, as the currency that one holds is able to meet the needs of the entities and individuals that make up the economy--as it is repeatedly accepted in exchange for sale of goods and services--the demand grows in this way, and as a factor of the ratio of the quantity of coin to the growth of the economy. Given a fixed quantity of coin, this ratio and thus each individual coin's price in goods and services will always gradually increase. To create a stable *value* for the coins, then, takes money supply. When discussing coins, this is physical minting in a coin foundry and coin press.
What else this implies, and adding to this, the greenback effect of the value imparted to a national currency by creating a tax liability that creates an artificial demand for the coins (demand for the currency so as to be able to pay the tax): is that neither backing paper money with bullion, nor being especially fastidious about a particular metallic type or content or ratio is necessary; these two effects, the "qualitative"artificial value imparted by the tax liability and economic demand building upon that foundation, and the quantitative value imparted by the growth of the economy--these two are enough, with on the supply side, proper minting of enough currency supply--by themselves to ensure a stable currency value. So if a metallic system is incorporated, for instance, in the paper currency, no systematic purchases into an ever growing reserve of bullion as currency is issued--no such system is necessary, but only preparation for the variability surrounding actual expected redemption in bullion--or coin, if a purely coin-based system. A mountain of bullion passively "backing" the currency plays absolutely no useful role, it seems to me.