FX Intervention and Achieving Central Bank Ends
Posted by Colin Lambert. Last updated: November 12, 2022
In its latest Expected Cost Monitor report, BestX carries a feature on the recent Japanese interventions to strengthen the yen, noting that the “gargantuan” sell order on October 21 “drained liquidity from the market and induced higher volatility and wider spread”.
The paper goes on to note that a “well-designed pre-trade TCA may help the central bank to better strategize execution plans and minimise market impact cost and maximise liquidity”, which raises an interesting question, doesn’t the Bank of Japan want maximise market impact?
At face value the question would appear to have a simple answer – yes, however while the second bout of intervention from the Japanese authorities has been followed by a decline in USD/JPY, it could be argued this is more the result of a wider dollar decline than any lasting impact from the central bank’s dollar selling. The BoJ spent over $20 billion in late September intervening to push the pair down and while it did have an immediate market impact, pretty quickly the market reverted and the market went substantially higher – thus prompting the October 21 event.
At a high level the central bank does not care about its market impact or the available liquidity – the core reason for the intervention is to shock markets into reverse – but could pre-trade TCA and a more thought-out strategy achieve better, lasting results?
The BestX paper observes one of the biggest challenges facing the Japanese authorities is the interest rate differential between Japan and just about all of its major trading partners and actually benchmarks USD/JPY to this rising differential. The chart (below) suggests that rather than being “irrational” or “extreme” as some Japanese government officials are wont to label FX moves, the market has priced in interest rate fundamentals rather well.
This means that for all its efforts to push USD/JPY lower, if the dollar bull run resumes – something that is not out of the question – Japan will once again be facing levels above 150.
There is a question of whether “old school” intervention – where the market is overwhelmed and liquidity is drained – is appropriate in today’s highly electronic market?
As everyone saw with the Swiss National Bank debacle over EUR/CHF, central banks do need to be careful in how they act in FX markets, and there is a question of whether “old school” intervention – where the market is overwhelmed and liquidity is drained – is appropriate in today’s highly electronic market? After all, the market has enough examples of how quickly a move can develop “flash” characteristics and threaten the efficient functioning of the market. That the BoJ’s intervention did not do so is probably thanks to the time of day it intervened, but should a central bank be creating the very kind of market upheaval and volatility that Japan’s Ministry of Finance in particular decries?
It could be argued that, as the BestX paper observes, a different approach could be taken. There is little doubt that intervention requires an initial shock, but perhaps what is needed is this shock followed by sustained pressure? In the instance of the latest Japanese intervention, the sale of perhaps $5-10 billion to get the market lower and then an algo working an offer over the next few hours? Not only would this have the desired effect, it would reduce the chances for a reversion, as seen in September.
The motivation for a trade is a big factor in any execution and in the case of intervention it is very much to move the market level substantially. Given the history of reversions – effectively in September a great number of traders quietly said a ‘thank you’ to the BoJ for giving them five big figures’ worth of better buying level – would it be better for the central bank to be more strategic in its thinking?
Either way, while it may not be obvious that the central bank wants a more subtle approach, there surely has to be value in the suggestion that it uses pre-trade TCA to inform its actions?