While the eurodollar system has attained the full functions of a banking system in parallel to (and in many ways superseding) the onshore dollar version, at its core remains its primary purpose as a means to solve global payments. It is the idea of the dollar replacing gold and sterling, the reserve currency. The difference in recognizing the eurodollar as opposed to the dollar is that there is no currency in that reserve system. For the most part, especially in the early days of the eurodollar, that was increasingly the point.
It's often a difficult concept to grasp, but I think it illuminates the core behavior that we even see today. Reviewing the early days of the 1960s is especially helpful, I believe, in describing these basic functions and constructions. It is particularly poignant as there is increasingly apparent symmetry in the nature of the eurodollar as it arose, and now in how it decays. At both ends, the Fed displays, and acts upon, only confusion - leaving just eurodollar banks and bank functions.
Toward the end of 1963, for example, the Italian lira was again under great pressure, as it was subjected to what modern central banks attribute more exclusively to naked speculation. With devaluation predictions ripe, the Bank of Italy was into mostly defensive actions. Part of the reason for that later in 1963 was that Italian banks had been heavily borrowing in the relatively new eurodollar market to circumvent domestic Italian "tightening." Commercial banks could easily convert what were ostensibly ledger liabilities in dollars (not physical Federal Reserve Notes) into lira or any other currency.
That had the effect in the international accounting of almost hiding Italy's "capital outflows." For perhaps the first time, the virtual eurodollar ledger exchanges were undermining official accounting and the standing upon which so many currency regimes were based. It wasn't just Italy, either, as Japan had undertaken much the same with its commercial banks, but without the imminent danger of devaluation; Japanese banks were actively sourcing eurodollars as a means to fund expansion without disturbing the yen at all.
As near as the Federal Reserve could figure, with assistance from the Treasury Department, Italian banks had sourced about $150 million in the eurodollar market toward the end of 1962, which equalized what would have been about $150 million in "outflows." The Fed, always taking the wrong track with the eurodollar market, suggested this would be a short-term issue, and that a reciprocal $150 million swap with either FRBNY or the Treasury would be sufficient when the "outflow" was eventually revealed.
By the following October, the eurodollar activity of Italian commercials had not ceased, so the Bank of Italy was forced to begin warning against further eurodollar financing. In anticipation of that warning becoming something of a soft prohibition, the FOMC was prepared to petition the Treasury Department to begin buying lira on the "open" market through the Exchange Stabilization Fund. FOMC officials were increasingly keen on this kind of action, because they believed less borrowing of eurodollars via Italy would lead to lower eurodollar rates, and thus help alleviate what they still saw as the US balance of payment deficit (it was, again, the incorrect interpretation of what eurodollars actually were and what the eurodollar market would become).
Later in October 1963, the Bank of Italy did make a "firm request" of Italian banks, which set in motion a series of transactions involving far more than lira - including, eventually, the dollar, the sterling, the yen and especially francs (Swiss, not French). Because "speculative" flow about the lira had become more intense throughout 1963, lira funds would often find themselves increasingly in Switzerland (a practice that continues to this day, no matter how much "floating" currencies). To counterbalance, the Bank of Switzerland was "forced" to buy up dollars (for reasons noted below) and send them off to various other nationalities and supranationalities (such as the BIS) via various swap lines which were rapidly approaching preset ceilings.
In essence, then, the US felt that its currency was "financing" the Italian "capital" deficiency if via Switzerland and the BIS. The effects of these arrangements were not neutral, and in a manner far more than disguising the precarious financial position of Italy. In FOMC discussions about the matter on October 22, 1963, the question was asked as to why the Swiss were not "financing" the Italian deficit directly through their own currency means. In other words, there really wasn't any reason for the dollar to be involved at all in the second leg of the imbalance from Italy to Switzerland; after all, Italian "speculators" were not arriving at dollars from lira, but francs. The dollar only appeared in this leg because the Swiss National Bank was using dollars rather than francs - the dollar was just the medium or tool.
The SNB, for its part, informally rejected Fed overtures that perhaps the Swiss might be better served using their own currency to stabilize their own currency. It was the Swiss view that the eurodollar market had become far more liquid and flexible to achieve these large programs without as much disruption (efficiency). In their own words, as reported by FOMC transcripts, the SNB had already been using sterling to alternately finance the Italian deficit and, having run into liquidity issues there, switched to dollars to "save their ammunition."
The nature of these imbalances was further distorted by official US actions which had the effect of, through both US Treasury and Fed actions (swaps and Treasury borrowing in foreign and forward dollar sales), covering what were ostensibly eurodollar holdings on the private side and converting them to official holdings at exchange values guaranteed by the US agencies. In the first nine months of 1962, the FOMC estimated some $600 million in Western European "dollars" were thus "converted" via official channels, effectively reducing uncovered (likely originating in the eurodollar market) holdings by $450 million (these are large numbers). In addition, as part of the wholesale shift in currency along these esoteric lines, foreign central banks as well as commercial banks began to hold volumes of US Treasury securities, primarily at first bills, rather than actual reserves.
When the Swiss National Bank claimed to "save its ammunition," this was part of its equation, preferring instead to source dollars indirectly - using lira to "buy" dollars in eurodollars and other channels, and then sending those dollars via swaps to the Fed or US Treasury, and turning the franc problem into another dollar issue inseparable from the eurodollar - which the Fed was hoping would just go away. That would leave out any part of mobilizing specific reserves of UST's to "finance" Italy. And so began the first definitive links between "dollar" behavior and USTs - whether or not "selling USTs" is actually selling at all (and in some cases, forgoing either selling or buying). At the end of 1963, due to events partially related to Italy and Switzerland, Japanese banks found themselves "short" by some $200 million in denied eurodollar rollovers. The Bank of Japan responded by "selling" ~$200 million in US Treasury bills, which was then only reflected in NYC money markets despite origination all over the place and across many seemingly unrelated formats.
While this is all a complicated and tangled mess (and I have left out a few steps for the sake of time and words), in many ways that is the whole point, especially since at its origin there was nary a dollar to be had or found. The ledger creations of the eurodollar led to a long and winding chain of events and traded liabilities that had the effects of converting virtual numbers into US government-guaranteed balances, further promises under reciprocal central banks, and all of which greatly affected very real financing and funding arrangements both overseas and domestically.
If there is any difference between 2016 and 1963, it is that the removal of fixed exchange had only the effect of removing central banks, largely, from the center of all these activities - they still go on, but financed and derived from commercial bank functions and balance sheet resources exclusively, rather than the official channels to stabilize currency exchanges. In this latter format, central banks enter, as described earlier today, as only the "last resort," and still a poor one.
In one some of the first official coverage of the eurodollar market, nearly a decade after it first arose in the mid-1950s, the BIS in its Thirty-fourth Annual Report easily recognized the substance of the system through (understandably given the period) a very traditional frame of reference:
In addition, it is claimed that long chains of interbank transactions may result in excessive stretching of the periods for which ultimate borrowers obtain funds as compared with the maturities for which original depositors place them, partly because narrow interest rate spreads may push the banks to seek higher yields through longer-term lending
In fact, the "long chains of interbank transactions" had the practical effect of not just moving bank resources further out in maturities, but also to increase and find efficiency in many other ways and means - to the point that "funds" and "original depositors" no longer apply, leaving only those long interbank chains. A good part of the ability to do that related to the dollar (and now "dollar") as, again, often nothing more than a vehicle for payment claims bridging many currency forms; as in lira to francs via dollar (or, as had been the case in the Canadian crisis of 1962, from Canadian dollar to sterling, francs, both Swiss and French, and marks via the US dollar). Over time, because of the complications and stretching chains, the "need" for actual currency simply faded away, since all that was really wanted was this system of liabilities (basically matching numbers at the end of each day from bank to bank, currency to currency, eurodollar node to eurodollar node) of purely "banking" resources.
The greater the complexity, obviously, the less ability to observe and interpret. That was difficult already in 1963; it is nearly impossible in 2016. Eventually, however, as 1963 and further on toward the full termination of Bretton Woods, the results show up somewhere.