Different traders have different ways of determining a currency’s fair value price. Obviously, different expectations and different motivations are what make a market. While many traders have their own decision rules and proprietary systems, there are currency valuation models and ways of looking at the currency markets that are the most common. We’ll cover the basics of a few of these currency valuation models below.
First, what gives a currency value?
A currency has value if it functions as an effective medium of exchange and a store-hold of wealth. When we think about the US dollar, we know it has both.
Globally, we know that US dollars are 62 percent of foreign exchange reserves, 62 percent of international debt, 57 percent of global import invoicing, 43 percent of FX turnover, and 39 percent of global payments.
The euro is close in terms of global payments, but the US dollar is far and away the most used currency in the world. Generally, the country with the world’s highest national income has the world’s global reserve currency. These regimes tend to change every so often.
From the 1400s onward, we’ve seen the world’s reserve currency change from Portugal to Spain to the Netherlands to France to Great Britain to the US. Empires, and hence reserve currency status, don’t last perpetually.
A reserve currency or reserve asset is defined as a large quantity of currency – typically owned through sovereign debt denominated in that currency – that is maintained by central banks and other large institutional investors, such as standard commercial banks and non-bank institutions, such as hedge funds.
Central banks will primarily use them as reserves, which can help influence their own domestic exchange rate. For example, selling reserves to buy their own currency will help it appreciate relative to the currency being sold. Large investors will invest in currencies to speculate – i.e., bet on their price movements – or to hold as currency hedges.
Different currencies do well in different environments. For example, Australia’s economy still materially depends on commodity exports, which increase in demand when the economy does well. So, the Australian dollar (AUD) is very pro-cyclical. When the economy does well, the AUD generally does well. When there are “risk off” episodes, the AUD typically sells off.
On the other hand, a currency like the Japanese yen (JPY) does well in risk off periods. Japan is a creditor country. In other words, it has what’s called a positive net international investment position (NIIP), or net external assets higher than net external liabilities. It lends more to the rest of the world than it borrows. During risk off periods in the market, Japan can pull back their assets abroad to be defensive, boosting the value of their currency. They also don’t have dollar debt related issues, or much debt denominated in a foreign currency (which can be dangerous because you have limited ability to control these liabilities).
The JPY is also a common funding currency because its interest rate is slightly negative. Traders like to borrow in yen to buy higher-yielding assets, commonly known as carry. If you borrow at zero and invest in something that gives you 5 percent yield, you make money off that spread. You would also need to ideally hedge out the currency risk. If, for example, you’re borrowing yen in the cash market, you could hedge that out by buying JPY contracts in the futures market (long JPY relative to your domestic currency) to offset your currency risk.
Of the seven main reserve currencies, this is one interpretation of how each one reacts based on what’s going on in the market and economy:
So, if you’re an Australian-based trader, you might consider that if your domestic currency is as pro-cyclical as it is, it might be beneficial to own smaller amount of yen and US dollars to offset some of that exposure.
With recent trade headlines, here are the measured currency sensitivities to world trade growth. Latin American currencies are more pro-cyclically sensitive, as are currencies that rely heavily on commodity exports, while the JPY is counter-cyclical.
There are also alternative currencies, or currency mediums and/or wealth store-holds that are not government-backed systems.
Throughout history, commodities have been used as a means of payment and to store wealth. Most commonly, this has been gold and, to a lesser extent, silver. This helps to theoretically stabilize a currency’s value as gold and silver are not subject to large swings in their demand.
Oil could also in some form be a currency, as an asset that will always in some form have value, though neither it nor precious metals are very effective as a transactional medium. Oil is much less effective than gold because of frequent material changes in its demand. Tying a currency to a commodity-based system can also be excessively onerous in that it limits the amount of money and credit that can be created.
Gold’s value mirrors currency and reserves in circulation globally relative to the global gold supply. As money is depreciated, gold’s value tends to go up in conjunction. For example, below we can see its value in relation to the amount of negative yielding debt.
Central banks trust gold as a source of reserves and large institutional investors trust it as a hedge against the depreciation of fiat currency.
Cryptocurrencies eventually have potential in a variety of ways. But they have a very long way to go to be accepted as viable stores of reserves by central banks or as hedges for large institutional investors.
Right now, cryptocurrency markets are heavily tied to speculative activity and not toward legitimate value creation purposes.
Currency Valuation Models
We’ll cover what are broadly the four most common currency valuation models:
1) Real effective exchange rate (REER)
2) Purchase price parity (PPP)
3) Behavioral equilibrium exchange rate (BEER)
4) Fundamental Equilibrium Exchange Rate (FEER)
Real effective exchange rate (REER)
The real effective exchange rate (REER) is determined as the weighted average of a country’s currency relative to basket of other currencies. The weights are a function of the relative trade balance of a country’s currency against each country in the basket.
Formulaically, REER can be expressed as:
REER = (E(a) x P(a)*/P) ~ (E(b) x P(b)*/P) ~ (E(c) x P(c)*/P)
E = nominal exchange rate
P*/P = ratio of the price levels
~ = “proportional to”
Another way to express it:
REER = ER^a x ER^b X ER^c x … x 100
ER = exchange rate of different individual countries;
The exponents (a, b, c, etc.) represent the trade allocation. For example, if a country does 30 percent of its trade with a particular country, the exponent would be 0.30.
A country’s REER can be found by taking the bilateral exchange rates between itself and all of its trading partners, then weighting by the trade allocation associated with each country and multiplying by 100 to form an index.
One of the implications of this formula is that if there is a real depreciation in a currency, net exports will increase. This makes sense because if a currency depreciates in inflation-adjusted terms, this will make a countries goods and services cheap in comparison. This boosts demand for them and hence exports will rise, holding all else equal.
Conversely, when a real exchange rate rises, this makes its goods and services expensive in comparison, reducing their demand holding all else equal, causing net exports to fall.
This is why many countries who use an export model to grow their economy will often want to see their currency depreciate. Those focused on a consumption model of growth will want to see a stronger or at least a stable currency, as their currency can go further (i.e., buy more goods and services on the global market).
This also means that certain trade relationships have more of an influence on an exchange rate than others.
For example, the US has a larger trade relationship with the EU – or more precisely, countries that use the euro – than it does Brazil. When the euro weakens against the USD, EU exports to the US become less expensive because the USD can buy more EUR on a unit for unit basis. To buy EU-based exports, US buyers (consumers, businesses, governments) need to convert their USD into EUR.
A move in the nominal EUR/USD exchange rate would have more of an influence in the USD- and EUR-based REER models than the weighting of the Brazilian real exchange rate because Brazil is much less of a trading partner by capital volume.
The dollar index (DX, published by ICE) is a weighted basket of exchange rates.
Currently, the index has the following approximate percentage weightings:
- Euro (EUR), 58 percent
- Japanese yen (JPY), 14 percent
- Pound sterling (GBP), 12 percent
- Canadian dollar (CAD), 9 percent
- Swedish krona (SEK), 4 percent
- Swiss franc (CHF), 3 percent
Purchasing Power Parity (PPP)
To compare economic productivity and living standards between countries, some macroeconomists will look at a metric called purchasing power parity (PPP).
PPP is a way that looks at the relative valuation of different currencies through a comparison of the pricing between different countries. The pricing comparison could be a specific good or a basket of different goods.
Mathematically, it can be represented as:
E = Pa / Pb
E = exchange rate between the two countries
Pa = price of the good in country A
Pb = price of the good in country B
Based on this idea, two currencies are in equilibrium with each other when the same good is priced the same in both countries taking into account the relative exchange rates.
To gain a meaningful understanding of the pricing differentials between countries, a representative basket of goods and services must be used, such as one that would represent the relative spending weights of all buyers in the economy.
This requires an enormous amount of data to be collected. To help with this process, in 1968, the United Nations and University of Pennsylvania developed a partnership, termed the International Comparison Program (ICP), to help facilitate this process.
The World Bank, on an every-three-years basis, will release a report comparing various countries in both US dollars and PPP terms. The Organization for Economic Cooperation and Development (OECD) and International Monetary Fund (IMF) will use PPP metrics to help generate economic forecasts and extend policy recommendations. The reports of these organizations can, in turn, impact financial markets.
Many currency traders will also use PPP metrics to help them find undervalued or overvalued currencies and inform trade recommendations. Currency fluctuations can also influence the returns of foreign investments in financial securities. A bond yielding 10 percent in a foreign currency that sees a corresponding decline of 10 percent in the currency would generate no actual return.
Some macro-accounting measures will adjust GDP for PPP. This helps to convert nominal GDP into a figure that helps to better compare between countries with different currencies.
For example, let’s say a pair of shoes costs $100 in the US and £80 in the UK. To make this comparison fair, we need to account for the exchange rate. Say the GBP/USD currency pair is trading at 1.30. This mean the shoes in the UK cost the equivalent of $104.
The PPP between the two countries, only taking into account this pair of shoes, would be 104/100, or 1.04.
This means that if a UK consumer wanted to buy the shoes more cheaply, they could buy them in the US for $4 cheaper. This would convert pounds to USD, leading to a rise in the USD relative to the GBP, holding all else equal.
What are the flaws in PPP in practice?
Before using the concept of PPP in your trading, it’s important to understand its flaws. PPP will be more or less valid depending on the situation.
Trade frictions and transit costs
In our shoes example between the US and UK, a UK consumer may want to save $4 per pair and just buy them in the US. That may not be feasible due to matters like shipping and transit costs. There may also be import duties that cause imported goods to trade more expensively in one country relative to another. The free flow of trade activity will help support the validity of PPP; restricted activity will impair its relevance.
There is an associated time delay, which represents another cost (“cost of convenience”). Many will be willing to “pay” that $4 to receive an item almost immediately relative to having to wait for an elongated overseas shipping process.
Some goods are also not tradable. Electric power is produced and sold domestically, as is hairdressing. Perishable items, such as some forms of food, can also not be feasibly traded.
In such indices as the Big Mac Index, input costs into the item, such as labor costs, maintenance, price of supplies, and so on, are not always traded between one country and another. These costs are not likely to run in parity with each other, especially not at the international scale.
Goods can be priced higher or lower in a country due to imperfect market competition. If a company has a strong competitive advantage in a certain market due to strong market power, this may keep prices of goods higher or lower. Because of a lack of perfect competition, expecting an eventual equilibrium in the prices of the goods, or a change in the exchange rate, is implausible.
Taxes are another force that causes changes to capital flows. Consumption and sales taxes, such as value-added tax (VAT) can cause prices to be naturally higher in one country relative to another.
Goods are not always of the same quality. Even seemingly comparable items (e.g., a Big Mac) can be of different quality between countries.
Price level measurements
Different countries have different baskets of goods that are representative of the consumption patterns within their economies. They will accordingly measure inflation rates differently.
The Big Mac Index
A popular form of the PPP is the Big Mac Index, introduced in 1986 in the publication The Economist.
Though largely introduced tongue-in-cheek – you can’t realistically expect the price of a particular food item alone to feed into an accurate representation of what currencies are worth what – the idea is to take the price, in local currency, of a Big Mac in one country and divide in by the price of one in another.
Here was the Big Mac Index from July 2018, adjusted for GDP per person (i.e., McDonald’s can charge people in richer nations more for a hamburger than they can in poorer nations):
Here are the six cheapest and six most expensive, and also the “fastest earned” and “slowest earned”. In other words, how long would the average person have to work in these countries to buy one:
The Balassa-Samuelson effect works to explain why consumer prices are structurally higher in more developed countries relative to less developed countries. It is suggested that this is due to greater variation in the levels of productivity in the tradables (relative to the non-tradables) sector.
In turn, this can explain the large difference in the price of services and wages between countries, and also differences between currency exchange rates and what PPP would suggest.
The implication is that currencies that represent countries with higher productivity will seem to be undervalued in relative terms and that this effect will appear to increase when incomes increase.
According to the “law of one price”, goods that are purely tradable should not differ significantly in price based on location. In other words, the price of socks in India should theoretically be the same as those in the US despite the large gap in GDP per capita. Both will be motivated to buy from the lowest cost producer.
On the other hand, most forms of services (e.g., gardening, medical care, dentistry) must be obtained locally. This makes their prices very specific to the location. Moreover, some goods, such as heavy furniture, have high transit costs – and could incur import duties – making the prices of these types of tradables deviate between markets as well.
The Penn effect stipulates that these higher price levels will structurally flow in one direction – namely, that countries with higher incomes will see systematically higher prices of services and goods with low value-to-weight ratios (e.g., furniture, heavy appliances, heavy machinery).
The Balassa-Samuelson effects also suggests that a wage increase in the tradable goods sector will also tend to lead to wage increases in the service (non-tradable) sector of the economy. Wage increases typically run higher in emerging markets where they have more catching up to do technologically, and thus see higher productivity rates. This increase also leads to structurally higher inflation rates in emerging economies relative to more developed economies.
Income and wage differential is ultimately predominantly an effect of the differing productivity between workers. Sectors that have low productivity advances are also the ones mostly involved in non-tradable goods, such as lawn care services. This needs to be true or else this type of work would be offshored (and obviously certain types of work cannot be feasibly offshored).
Certain jobs are less sensitive to advances in productivity relative to others. For example, a seamstress in New York is generally no more productive that one in Kiev. But these types of jobs need to be completed locally. So even though workers in New York taken collectively are more productive than workers in Kiev, the equalization of local wage levels causes New York seamstresses to be paid more, on average, than seamstresses in Ukraine.
In other words, local goods (non-tradables) have different prices depending on the jurisdiction and will tend to be more expensive in richer countries and cheaper in poorer countries. Tradables will be more or less the same price in any country, though this is dependent on transit costs, trade barriers, taxation, and other such factors.
This also means that the tradables sector is more influential on a country’s exchange rate relative to non-tradables. This is because when one country wants to buy from another country, they need to swap currency, and this influences the supply and demand. If a US buyer wants to buy something from Japan, he must convert his dollars into yen. This boosts the yen relative to the USD.
Behavioral Equilibrium Exchange Rate (BEER)
The BEER approach attempts to measure the exchange rate misalignment between any given two currencies based on transitory factors, chance disturbances, and current economic fundamentals in relation to their sustainable levels. The BEER approach is often employed using econometric applications and may be used to explain cyclical changes in the currency.
The choice of variables for the BEER approach is discretionary, based on beliefs of what impacts an exchange rate and the data that’s available. This can include the following:
– monetary policy and its future likely path / nominal and real interest rate differentials
– terms of trade (price of imports and exports between countries)
– national savings and savings strategies
– productivity differentials
– debt and equity capital stocks (and the risk premia between them)
– demographics and their net effects on savings and other forms of economic behavior
– net foreign assets relative to net foreign liabilities
– fiscal policy and its future likely path
– levels of foreign exchange reserves
– capital account policy
– net external debt as a percentage of output
– prices of traded and non-traded goods
– tariffs, import duties, non-tariff barriers, and other such macroeconomic factors
Some might also consider such qualitative factors such as the result of a political election or action (e.g., Brexit news), but these are largely influential through expected changes in macroeconomic variables as a consequence of these events.
For example, the election of Donald Trump in the US impacted how traders perceived where personal and corporate tax rates would go. This impacts capital flows.
If tax rates are lowered, for example, the expectation is for more capital to flow into the US. It could also increase consumers’ take-home pay and increase the national savings rate. If lower corporate tax rates also incentivize companies to invest more, then this would have positive productivity outcomes. It could also cause the fiscal deficit to widen (if the extra income produced doesn’t offset the spending and result in higher tax receipts) and lead to the need to sell an unsustainable amount of bonds (a promise to deliver currency over a period of time) externally.
Linked here is one such example of a BEER econometric approach to determining fair values in different exchange rates.
Fundamental Equilibrium Exchange Rate (FEER)
A fundamental equilibrium exchange rate (FEER) is similar to how the term is used with respect to the valuation of other asset classes. What is the fair value exchange rate on the basis of existing policies? In the realm of currencies, the fundamental value should be one that is expected to generate a current account of some surplus or deficit that is equal to the country’s underlying capital flow. This makes the assumption that the country is looking to pursue internal balance and not restricting trade or capital flows to keep its balance of payments at a certain level.
For example, if a country wants to peg its exchange rate, it either needs to forgo having an independent monetary policy (if it wants to allow capital inflows and outflows to move as needed) or restrict its capital account if it wants to continue to have autonomy over its monetary policy. (More on this dynamic, commonly called a trilemma, is explained here.)
Because productivity and overall output tends to grow over time, this increases countries’ holding of reserves. The secular growth rate in reserve holdings also gives an indication of the amount of foreign capital that could be available to finance a current account deficit for countries that have them (e.g., the US, UK, Canada, Australia, India).
Because capital is mobile, knowing what a country’s underlying capital flow is exactly is hard to know. It would be dangerous for any country to assume that any amount of fiscal or current account deficit could always be financed through external capital (selling debt to foreigners). The would make the concept of a fundamental equilibrium exchange rate moot.
Borrowing extensively (a form of capital inflows) and lending extensively (a form of capital outflows) can be economically dangerous and unproductive.
A FEER must be defined in real (i.e., inflation-adjusted) terms. A country that’s undergoing an inflation rate that is five percent higher than that of other countries means that its currency will need to depreciate by five percent in order to restore the same competitive position as before. Namely, the five percent depreciation will give consumers the same set of choices as they had previously and producers needing to sell their goods will have their competitiveness restored internationally.
The exchange rate considered is also not a bilateral rate in the sense of “the EUR/USD exchange rate is 1.10” but rather an effective rate. In other words, other countries are taken into consideration and weighted based upon their allocation in the foreign trade of the country in question. This then feeds into the estimate of the overall exchange rate that measures an overall competitive position.
It is misleading to measure a country’s “exchange rate” as one that takes into account only the currency of a single trade partner, assuming that the country’s trade is diversified among several countries.
Complications in finding the FEER can be difficult for certain countries that depend heavily on the exportation of a certain good such as oil and their particular national savings strategy.
For example, Norway is an oil-exporting country that converts its savings into ownership of mostly foreign assets. Another oil-exporting country, such as Ecuador, does not save anywhere near as aggressively. Therefore, we can expect Norway to be less sensitive to an oil price decline relative to Ecuador.
Under normal circumstances, oil prices are going to be a significant influence on the exchange rates of these particular countries. For non-oil exporters, a country’s sustainable current account level will be the most influential long-run structural factor that impacts the equilibrium exchange rate.
It is commonly, though somewhat arbitrarily, believed that countries should not run current account deficits of more than around 3 percent of GDP. To fund this deficit, countries will need to finance it through debt. For the same reason, it is believed that because deficits and surpluses are a zero-sum game, this rule should symmetrically apply to current account surplus countries. This is part of why purposeful currency depreciation (which makes exports relatively cheaper internationally and helps build a more positive trade balance) is frowned upon.
Some macroeconomists suggest that a +/- 3 percent surplus/deficit situation can be exceeded so long as surplus countries are not increasing their net foreign assets relative to GDP and deficit countries are not decreasing their net foreign assets relative to GDP.
How did this “3 percent rule” come about?
In the past, academics and market practitioners identified ~40 percent as an approximate level of external debt to GDP that should not be exceeded. Empirically, this tended to increase a country’s vulnerability to default. If emerging market growth rates are 4 percent in real terms and 7 percent in nominal foreign currency terms (and 6 percent in nominal USD terms), this 40 percent multiplied by 6-7 percent produces a value of 2.4 to 2.8 percent. And this assumes that the deficit is financed in domestic currency where the rates and payments can be controlled, unlike if it’s denominated in a foreign currency.
This 2.4-2.8 percent represents the theoretically sustainable deficit level. Because growth and inflation were somewhat higher in the past, this led to a higher equilibrium estimate of the sustainable current account deficit. It differs on a country by country basis. Some countries have structurally higher growth and inflation rates and will also depend on the country’s net foreign asset position. If a country earns more on foreign assets relative to what it pays out on liabilities, this extra income generated can support a higher deficit.
Reserve currency countries also enjoy the advantage of being able to sell their debt to the rest of the world relative to a non-reserve currency country. Though the US has a balance of payments issues with a fiscal and current account deficit, the US dollar is the most widely used currency in a variety of functions, so there is a large amount of foreign demand for it.
Thus, the US can keep its deficits structurally higher relative to a country that has economic and political instability, less robust institutions, higher levels of corruption, less support for commercialism and innovation, less internal investment, inadequate capital markets development, and so forth.
Nonetheless, the US is a country that is likely to only growth at some 3.5 percent nominally (real output plus inflation) going forward, with about 1.7 percent real output and about 1.8 percent inflation. The breakdown in real output can be found below:
With the world’s reserve currency, the US enjoys the advantage of being able to run a higher current account deficit simply because it’s able to sell a lot of its debt to other countries who want it. The US effectively enjoys an income effect from this and is why it’s GDP per capita is higher than similar countries, such as Germany, Japan, and the UK. Since the end of World War II, the real GDP per capita of Germany, Japan, and the UK have stabilized at around 70-80 percent of the US’s. The dollar’s influence has been a big part of this. Namely, it can enjoy higher living standards, which makes having a reserve currency important.
But this amount of debt issuance has its limits. The US has an external debt to GDP ratio of about 45 percent. If nominal growth is only 3.5 long-term and we assume that the US will have trouble selling even more of its debt to foreigners who are already have USD reserves at ~60 percent of all reserves, this means current account deficits of more than 1.6 percent will be hard to sustain.
The more the US stays loose with its spending patterns, the more it endangers its status as having the world’s top reserve currency. In the near-term it’s not an issue. The US has the highest national income, which typically conveys top reserve currency status. But it’s an issue over the next decade-plus.
Currency valuation models summary
In total, we have four main currency valuation models that are broadly applied:
1) Real effective exchange rate (REER)
2) Purchase price parity (PPP)
3) Behavioral equilibrium exchange rate (BEER)
4) Fundamental Equilibrium Exchange Rate (FEER)
The REER is defined as the weighted average of a country’s currency relative to a basket of other currencies. A country or jurisdiction’s REER can be calculated by taking the bilateral exchange rates between itself and its trade partners and then weighing each exchange rate by the trade balance (in percentage terms).
PPP suggests that exchange rates should equilibrate over time such that tradable goods between countries become equal in price. For example, if a shirt costs $10 in England and $15 in the US, it suggests that US consumers will want to buy the shirt from England (thus converting US dollars into British pounds), causing an appreciation in the GBP and depreciation in the USD.
Problems with PPP include the distorting influence of structural and idiosyncratic factors between countries, such as trade frictions and transit costs, non-tradable input costs, imperfect competition, tax disparities, goods quality, and price level measurements.
The BEER uses an econometric approach to measure the exchange rate misalignment between different currencies based on temporary influences, random events, and current measurements of economic fundamentals in relation to long-run sustainable levels. Often it is used to explain cyclical influences in the currency.
For example, in the US, each April the US Treasury sees an influx of revenue. This pulls liquidity (i.e., cash) out of the economy as people and businesses pay taxes. Less liquidity in circulation pulls up the value of the USD taking everything altogether. An econometric approach could work to understand this effect.
The FEER uses the concept of a fundamental equilibrium value. For example, if a country is running a balance of payments deficit (combined current account and fiscal account is negative), can it plausibly fund that by issuing debt to foreigners?
Reserve currency countries that issue debt in their own currency where there is strong external demand for it will be able to do a better job of this than emerging market countries. Emerging market countries that issue debt in foreign currency (generally because it has lower interest rates and it’s more stable in value than their own domestic currency) will run into issues if their currency depreciates against the currency in which many of their liabilities are denominated.
Currencies with structural balance of payments surpluses often see their currencies appreciate, as it indicates a surplus of demand for the currency on the international market.
All of these models assume floating exchange rates.
When there is a fixed exchange rate, the central bank or broader monetary largely controls the exchange rate by buying and selling on the open market. Some other currency regimes are what you would broadly call “managed”. Namely, they are neither floating nor fixed but kept within a desired band for some period. The Chinese renminbi (also known as the yuan) is one example.
In terms of policy choices, a country will need to fix its exchange rate by either controlling capital flows in and out of the country (through the banks, non-bank financial institutions, and other conduits) or by forgoing the use of an independent monetary policy.
For example, if a country wants to peg their currency against the US dollar and don’t control their flows (because they don’t want to or because they don’t believe they realistically can), then their own monetary policy will be dictated by the US Federal Reserve. This keeps the interest rate differential relatively constant. When one currency earns more or less relative to another, this makes the currency more or less attractive to hold.
A fixed exchange rate can also be feasible when capital flows between countries are small to the point where managing an exchange rate is relatively easy without flows subjecting it to large changes in demand. This is part of what made the Bretton Woods monetary system effective from 1944 to 1971.
Even though there are various fixed exchange rates throughout the world – e.g., Hong Kong dollar relative to the US dollar (USD/HKD) – these currency pairs are still tradable. Just don’t expect them to move much unless the peg is broken.
In the end, all pegged systems that are inconsistent with the fundamentals of the currency will eventually fail.