The Currency Exchange Markets

This is a qualitative difference from the 1930's Great Depression: the trading triangles are now "rigid" rather than "floppy". I haven't thought out whether this should amplify or mitigate the effects.

However, considering the provocation to violence in Indonesia (ethnic Moslem vs. ethnic Chinese, ethnic Christian, or police stations; particularly since Ramadan 1998) and South Korea (labor unions directly clashing with military and/or police; particularly since January 1999), I would speculate that the net effect of "rigidizing" the currency triangles is to amplify the effects.

Intercurrency trade

• If the price is in the buyer's currency, this creates an automatic supply of buyer's currency.
• If the price is in the seller's currency, this creates an automatic demand for the seller's currency. [This is often illegal; for instance, it is illegal to price goods in U.S. dollars in Brazil. At least, this was reported in CNN Custom shortly after Brazil's real devaluation.]

• Japanese exports to the U.S.A. [Automatic supply of U.S. dollars]
• U.S.A. exports to Japan [Automatic supply of Japanese yen]

Let's assume that the BoJ gets its target rate (as of somewhen in May 1999) of 120 yen/U.S. dollar, and that this is relatively stable. Let's furthermore fantasize that the trade surplus with the U.S. is, say, 1.0 trillion yen/month. Supposing the exchanges are open 24 hours/day, let's pretend that we're in a 31-day month, so the trade surplus for a given day is about 48 billion yen. [Both "trillion" and "billion" are American: 12 and 9 zeros, respectively.] [We will also do a what-if the trade surplus was 1.5 trillion yen/month. These figures are from the March 1999 trade surplus figures reported in May 1999. The what-if has a trade surplus of about 72 billion yen for the day in question.]

Now, realize that this figure is resulting from:

• "U.S. sales in Japanese yen"-["Japanese sales in U.S. dollars"*"Exchange rate"]=48 billion yen

• ["U.S. sales in Japanese yen"+"other sources of yen"]/["Japanese sales in U.S. dollars"+"other sources of U.S. dollars"]=120 yen/dollar="Exchange rate"

• If there were no other sources of either yen or U.S. dollars, then the left-hand-side would be zero, which is not allowed.
• If we substitute in the right-hand side of the exchange rate, we get
• "U.S. sales in Japanese yen"-["Japanese sales in U.S. dollars"*["U.S. sales in Japanese yen"+"other sources of yen"]/["Japanese sales in U.S. dollars"+"other sources of U.S. dollars"]]=48 billion yen, which rewrites as
• "U.S. sales in Japanese yen"*["Japanese sales in U.S. dollars"+"other sources of U.S. dollars"]-["Japanese sales in U.S. dollars"*["U.S. sales in Japanese yen"+"other sources of yen"]]=(48 billion yen)* ["Japanese sales in U.S. dollars"+"other sources of U.S. dollars"] [I]
• [The formal units of the above rewrite are yen*dollar, which is (as far as I know) economically meaningless.] Fantasize that in the revised projection (72 billion yen), that the only variable is (Japanese sales in U.S. dollars). This obviously tends to strengthen the yen. Note that Le Chatlier's Principle (from chemistry) probably applies here, in which case the actual shift will be in the same direction, but less than, the directly computed shift. EXERCISE: compute the interval within which the above shift must fall.

Intercurrency loans

Observe that when an intercurrency loan is disbursed, it creates an automatic supply of the lender's currency (which generally is diffused over days or months).

The problem is when an intercurrency loan is repaid. Intercurrency loan repayments create an automatic demand for the lender's currency. [Now, we see why it is often illegal to price imports in the exporter's currency...] If automatic demand for the lender's currency exceeds automatic supply (from the trade surplus and other sources), there is a risk that the borrower's currency will have a 'desperation boost' in supply. This 'desperation boost', if not matched by a supply boost in the lender's currency, will result in a shortfall of actual demand of the lender's currency relative to automatic demand. Mass loan defaults follow.

Malaysia had the unusual situation that a major Malaysian oil company did its accounting in U.S. dollars. This allowed Malaysia to defend its ringgit with the oil company's income stream, rather than Malaysia's central bank reserves. This was critical in Malaysia's preliminary defense against the Crash, buying time for Marathir to construct the Malaysian Firewall.

This is obviously a key factor in the collapse of the Thai baht, the Indonesian rupiah, and the South Korean won. At the start of each of these currency collapses, the percentage of intercurrency [international] loans due in six months, out of the total, was 56%-57%. The Phillipines, whose currency did not collapse, but did waver, had a maximum percentage of intercurrency loans due in six months of 55%. Apparently, 54% did not cause symptoms.

Note that none of the four countries, above, is openly documented as having circumstances like Malaysia's. Also note that we seem to be dealing with a "threshold effect"; qualitatively, the cliff is somewhere between 54% and 56%, and the edge is near 55%.

George Fisher has pointed out to me that there is a historical analogy of the British trading houses [1600's?] to the current situation. "... so, there is nothing new under the sun." [Summary of Ecclesiastes 1:1..11, and the last part of Ecclesiastes 1:9, NASB]

• In the U.S., collectively, we do not have reasonable reserves of food. I have heard (without reasonable documentation) that U.S. supermarkets are planned with a turnover time of 3 days on basic staples. Certainly, the shelves at Walmart in Paola, KS can get very bare on Thursdays before restocking.
• I have also heard (with reasonable documentation) that only 1%-2% of the U.S.A.'s money supply is in cash.
• In 1981, it was estimated on TV that a 7.0 earthquake in the Los Angeles area would "disrupt" the U.S.A.'s banking system in 1 day (presumably by paralyzing a few hundred thousand checks?). This was before the advent of ATMs and debit cards (the Apple II was released in 1979!), or electronic funds transfer. Now, the disruption time could be mere hours. Or minutes. (How long does an EFT bank run take to shut down a bank? If anyone reading this knows about a well-documented account, please email references).
• Also, consider the major icestorm that leveled power lines south of Montreal. (Confirmation requested: January 1998 or January 1999?). The Canadian armed forces were called in on the eighth day without power -- in part, because in Montreal the food was about to run out. It took two months to restore power south of Montreal, and at least two weeks (again, confirmation?) to restore power in Montreal.

• I seriously advise against taking any sort of publicly visible radical preparation for any disaster -- it makes you a highly visible crime target during such an event, which is counterproductive. I do think that subtle changes in routine, applied for months or years, that are applicable to many different implausible disasters, are reasonable.
• In the U.S.A., the FEMA-recommended resources are definitely reasonable. The measures listed at Noah's Ark are closely related to the FEMA measures.

Return to the Crash page, or the main page.