This post is a fictionalised thought experiment. YOU and YOU ALONE are in charge of what happens in this story. There are trade-offs, choices, and consequences. YOU must use all of your B.S.D. talents and much of your intelligence. Good luck.
You’re the head of a trading desk, and you’ve felt a bit frustrated of late.
Your desk trades government securities issued by the world’s largest economy. This is definitely not America we’re talking about, so we’ll call them, um…Tresories.
Your main frustration is that you’re a primary dealer, which means you’re required to bid for a pro-rata share of each government auction. That can be a capital-intensive undertaking, and because regulations have made balance sheet space valuable in this alternate universe, you need to be careful about where you commit capital.
After the fiasco at Salomon Brothers in the 1990s, you know better than to flat-out cheat in your dealings with this government. But you still wonder — rather stupidly, perhaps — if you could stretch the rules to ease the pressure a bit and improve your returns.
Here’s an ultra-simplified description of how that type of auction works:
Let’s say the government has $25bn of two-year notes to sell, and they get $60bn of bids (since every auction is oversubscribed by design). An over-simplified book of bids might look like this:
$5bn of bids at 1.2 per cent
$10bn of bids at 1.25 per cent
$15bn of bids at 1.3 per cent
$15bn of bids at 1.35 per cent
$15bn of bids at 1.4 per cent
The auctioneer starts by allocating securities to bidders who offered the highest price (or lowest yield) to purchase them. So the banks, investors and corporations who offered to buy two-year notes at 1.2 per cent would end up with their $5bn of Tresories.
But those buyers wouldn’t end up with a 1.2-per-cent yield. Instead, the seller keeps going down in price (or up in yield) until there’s enough demand to cover the entire amount sold. And then everyone gets their bids filled at the clearing price, according to this handy piece from the New York Fed.
In this stylised example, the final clearing price is 1.3 per cent, and the auction allotment would be as follows:
$5bn of bids at 1.2 per cent — bidders get $5bn at a 1.3-per-cent yield
$10bn of bids at 1.25 per cent — bidders get $10bn at a 1.3-per-cent yield
$15bn of bids at 1.3 per cent — bidders get $10bn at a 1.3-per-cent yield (all of the bidders would get some Tresories, with their share of the $10bn determined by how much they wanted at this price)
$1bn of bids at 1.35 per cent $1bn of bids at 1.4 per cent
Buying the most securities is a tricky undertaking, since there’s a real risk of bidding too high. If you do, you could move the price up, which would make it tougher to “flip” the securities by selling them for a higher price in the secondary market.
But there are other ways to game this (hypothetical!) system, which does closely mirror the current US auction format.
You, our (purely fictional!) trader, have a few options:
A) Try to get the clearing price exactly right, and make your highest bids at that price to minimise risk.
B) Try to make sure you get paid in secondary markets.
C) Do nothing.
A) The Clearing Price Is Right
You’re a bright guy, and you employ a lot of other bright guys. So you figure that if you really put your minds to it, you could get the price on newly issued Tresoriesexactly right, and make sure that price is where you put in your highest bids.
Remember how, in the Dutch auction example above, the bidders who offered to buy at the clearing price didn’t get all of the securities they wanted? This is what happens to you. Your “hit rate” suffers.
But that’s not a bad thing for your bank! Those pesky capital requirements have made it less attractive to take on a lot of Tresory-auction-related risk. Of course, this purely hypothetical scenario doesn’t look much like what’s being addressed (in a confusing manner) in this New York Post article.
Really, does it count as “winning” to get a clearing price right and … not buy as many securities as your primary-dealer peers?
Maybe. But it doesn’t guarantee a profit, either! It just frees up precious balance sheet to be deployed elsewhere.
Still, everyone’s happy that you’ve minimised the capital used in Tresory auctions. And you’re getting lots of client trading volume every day, as investors prepare for Brexit and Fed rate hikes. Because this is still a market that trades in fractional pricing and high notional sizes, you do pretty darn well, though your desk isn’t as well-staffed as it used to be. (It’s not risk money, but who makes risk money these days, anyway?)
Your boss loves that your desk is now a dependable source of revenue. You get promoted. Congratulations!
Of course, the government is always bothering you about low hit rates and your prescient bids. Clearly, they just don’t understand your genius.
B) Secondary Market Shenanigans
You’re really unhappy about the fact that you’re required to take a bunch of risk by bidding at auctions. And these days, a significant number of buyers are bidding directly, so you can’t say for sure you’ll have a great idea of what the final prices will be.
But then there’s a when-issued market. And you do get to see your clients’ bids.
The when-issued market is where you pre-sell Tresories, basically. Big clients like index funds know they’re going to need them, so you guarantee delivery of the new security to them at a price determined before the auction. This means you have what amounts to a short position in the new security. It does not mean you’re doing anything wrong.
It does, however, mean that you have an inherent conflict of interest with the government, thanks to this partially-but-not-entirely-private auction structure. While the government wants to keep its auction price high (and yield low), you now benefit if the auction price is low (and the yield is high). You also want to push prices higher in the when-issued market, so you can collect the spread between the two.
If you were officially underwriting the auction, you’d be able to charge fees, and if the auction process were entirely public, you wouldn’t be required to bid for your pro-rata share. (The public auctions would likely fail sometimes, but probably without many negative repercussions, which is the way things work in Germany.) Yet you’re required to bid, but you don’t get fees, which gives you a strange quasi-underwriter status.
As a result of this status, you have some leeway with your desk’s conduct. For example, it’s entirely reasonable to charge a premium for when-issued securities. Why else would you offer them in the first place? Your traders can figure out reasonable prices for when-issued securities by looking at the client demand for when-issueds and their bids for the auction.
After a while, you get a little more confident. You don’t yell at your traders when they look at the order book during the auctions, which allows them to outbid clients by a slight margin and then flip the new securities into the secondary market. It also helps them make sure they aren’t bidding too high. The auctioneer visits every year to inspect your desk and practices, and there’s no rule against looking, so why would you ban it?
Your traders do well with this system. They don’t let auction prices get too high, and they don’t let when-issued prices get too low. Then, you realise that they’ve been chatting with traders at other primary dealers, to ensure they won’t be caught on the wrong side of unexpected demand from other banks’ clients.
This introduces problems. Serious problems.
From the traders’ perspective, they’re just acting like normal underwriters would, and engaging in price discovery before an offering. But according to a group of large investors who have filed a lawsuit, they would be colluding to rig auctions. That would get the Justice Department on your case, and all of a sudden you’d be the recipient of a subpoena asking about your when-issued trades. Good luck.
C) Keep Your Nose Clean
You care about your country and its borrowing costs, which is a good thing! Sure, you develop a bit of a reputation with the auctioneer as a gadfly — you’ve noticed what the other primary dealers are doing and complained — but they think of you as a harmless nuisance.
You work at the primary dealer for a few more years, avoid trouble, and then move to a money management firm. The work is less risky (and less remunerative) but more stable. Plus, it was getting a little bit depressing to watch your traders’ jobs at the bank get automated away.
In any event, you haven’t risked your job or the money of the bank’s shareholders. Bravo!