Transaction costs 101
Today has been an interesting day regarding market impact and how despite huge volumes a large order can potentially move markets. For many of you like my other "101s" this will be obvious but for others it probably new.
Every order has a transaction cost to get executed. It may be just selling on the bid side of the market. It may be selling on the bid side plus commission and it may be pushing the equilibrium price in order to move large size the last bit is the focus of this thread
In order to move large size transferring your risk to someone else you have to offer a concession. Essentially a discount to the equilibrium price when selling a large block. That discount encourages players who were perfectly fine with their pre existing positions to buy.
Their expectation is that some or all of the discount will be captured if they hold your block longer than you were willing to hold it. Market makers and other temporary liquidity providers are constantly seizing on this discounts. It's fairly competitive
That's the discount part. However MM's are wary of buying a large block at a discount if they expect you have another large block to sell right after they buy the first one. You also are wary of disclosing your complete hand (picture is the lingo). Also you may have alpha
Meaning those who take the opposite side of your trades on average lose money because when you sell you are usually right.For that reason information about who is selling and the actual size of the order is extremely valuable and any information slippage can increase the discount
So there are all sorts of ways to get prevent information slippage and to execute orders in ways to minimize the discount. Nonetheless even the best of the best will still see a discount if they want to move more volume than the market can take
For three decades I have helped clients at Salomon minimize transaction costs and for my time on the buy side always been concerned about market impact and cost. I am not alone!!!
Every serious money manager regardless of the type of manager is focused on trading with low market impact. Trading desks at institutional buy side clients are judged and compensated by their firms based on costs of trading. Needless to say the area is quite well studied
Most firms collect data of every order they do in every asset class and specific security. Data sciences then allows them to understand exactly what they should expect to pay in market impact for every security they trade in every size and order timing.
These multi dimensional curves can even be extrapolated for larger size then they have had experience in doing by using related experience in other securities where they have traded large size. Then every trade can be judged based on the ex anti estimate and the ex post result
That analysis is then incorporated in the data and the curves are iterated. Every big client of Salomon and every hedge fund I know of uses some sort of analysis of this type to trade. It's the price of admission to play in this business
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Us naive Americans dont think about currency returns as part of our portfolios as we have the biggest and for decades best place to invest in equities.
Every other global investor cares about currency returns at basic level for their investing
The basic idea for investors or all nationalities should be simple and obvious to all. But we Americans just haven't had to care. Maybe we still don't but at least we should be aware. This 101 will explain what is obvious to all non Americans and then show how it works
The goal of all investors is simple. We want to maximize the risk adjusted return of our investments in the currency we expect to spend in the future.
As Americans we want to maximize our USD returns
If we are Japanese we want to maximize our Yen returns
Money creation and credit creation in the private sector 101 part 2.
Role of Repo.
In the prior thread I outline credit creation which can happen without banks and money creation which requires banks.
I also hinted at bank reserves role as being one of grease to the
system and NOT necessary for bank money creation but necessary for interbank deposit shifts. I also didn't discuss base money creation from the Fed and won't be dealing with that in this thread either.
Here I will discuss the specific role of Repo in today's financial system
The big takeaway is it is one of many important and necessary means of credit creation AND it has no role in money creation unless a bank is a party to the transaction.
That will take some weedy mechanics to prove. But before we do that let's talk about the entire economy
Money creation and credit creation in the private sector 101
There has been a lot of focus on the repo market lately. I get it. It's an important part of the capital markets in the credit creation process. But its growth and contraction is part of the credit creation process
The repo market where transactions are between hedge funds and money market investors, and those who desire leverage for whatever purpose is an important market in the credit creation process BUT is not part of the money creation process UNLESS a commercial bank or the Fed is
A party to the transaction. Because this is largely misunderstood by even some plumbing experts it's worth it for me to write out my understanding (maybe im wrong which would be awesome so I can learn). So here I go.
I've been studying various versions of balance sheet expansions over my career. I'd classify them as
Japanese first failed effort
UK's version
U.S. Version 1
U.S. version 2
ECB version
Japanese all in version 2
They are all fairly different in approach. The big takeaway 🧵
The developing Fed version that most are excited about is most akin to the Japanese first failed effort.
Here's a rough summary of each
In 2001-2006 Japan the BOJ initiated QE. In their version they offered significant lending to the Japanese banking system for good collateral
The balance sheet doubled in size at a pace of 35 Tn yen per year. However of that 35tn only 5 was direct asset purchase and most of that was Japanese Tbills. This is very similar to the BTFP program from SVB time and the current SRF. It was also sorta similar to ECB LTRO
Why do repo rates change and what do they have to do with reserves. This is a super technical issue and there are better folks to follow on this topic than me but I'll give it a go.
Firstly what are the two sides of a repo transaction and why do they want to interact.
One side is a guy with a bank deposit he wants to earn interest on. The other is a guy who wants to borrow money overnight and has assets he owns that he is willing to provide as collateral to the loan. We can go down a level on each side but for now let's keep it simple.
Most repo transactions are done with UST as the collateral and most UST collatarel used is TBills but. UST's are also highly common collateral but do to the marked to market risk they offer less borrowing capacity per unit of notional (higher haircut)
Some thoughts on 10 year notes since Powell guided for a restart of the cutting cycle at Jackson Hole. Trying to answer what the bond market is saying
Nominal yields have fallen 33bp
Note yields are driven lower by
1)Falling real GDP expectations
2)Falling Inflation expectations 3) Falling "risk" of owning assets 4) Improving supply/demand balance vs expectations.
In attributing nominal yield changes to these 4 things unfortunately market prices don't
Easily demonstrate these things. For instance 3&4 are only able to be measured via a model which estimates risk premiums or the expected return over holding cash
Even Breakeven inflation and real TIPS yields have risk premium buried in there market yields. However we can try