WhitmanTrading

Investment Bank: Paid for the Deal

An investment bank arranges and finances transactions for institutions: advising on mergers, underwriting share and bond issues, making markets and publishing research. It is paid for the transaction happening rather than for how it turns out, and that single fact explains most of its behavior.

How it works

It is an intermediary between organizations that need capital and organizations that have it. No deposits, no current accounts, no retail customers — a different business from the bank where your salary lands, sharing only the word.

Four activities under one roof. Advising on mergers and acquisitions. Underwriting new share and bond issues. Trading — making markets and taking positions. And publishing research on companies and markets.

The fee arrives when the deal closes. Whether the merger creates value three years later, or the newly listed company trades well, does not change what was earned. That is not a scandal; it is the structure, and knowing it explains a great deal.

The activities conflict, and the industry knows it. A firm advising on a deal holds information its trading desk must not have, and research covering a company the bank wants as a client is under obvious pressure. Internal barriers exist precisely because the conflicts are structural rather than accidental.

What that means for a share price

In a flotation the company selling shares is the client. The bank’s job is to place the issue at a price that satisfies the seller and still clears. That price is a negotiated outcome, not an estimate of what the business is worth.

Analyst research is a product, and the reader is not paying for it. It is distributed to institutional clients and is genuinely useful for the detail it contains — industry structure, capacity, contract terms. The rating on the front is the least informative part of it.

A price target is not a forecast you can act on. It is a house view published on a schedule, revised after prices move as often as before, and it carries no risk rule of any kind.

In practice

Underwriting fees are deducted from the money raised. The company receives the issue proceeds net, which is one reason flotation pricing is negotiated as hard as it is.

Market-making means holding inventory and quoting both sides. The desk is not betting against you in any adversarial sense; it is managing a book, and your order is one input to that.

The business is cyclical in a specific way. Cheap borrowing produces mergers and flotations; expensive borrowing does not. Revenue follows the financing environment more than it follows the stock market.

Research reaches clients before it reaches the public. By the time a rating change is reported, the opening gap has usually already happened.

Nothing in a research note tells you where you are wrong. A target is a destination without an invalidation, which makes it unusable as a stop.

Acting on any of it has a price. Every round trip pays the spread and any commission, and by the time a rating change reaches a retail screen, what is left of its effect can be smaller than that toll.

Reading a research note usefully

Skip the rating and the target, and read the assumptions. The valuable content is the modeling — what the analyst assumed about volumes, pricing, margins and capital spending, and over what period.

Those assumptions are testable and the conclusion is not. You can disagree with a growth rate and see what it does to the answer. You cannot do anything with “buy”. A note whose assumptions are stated clearly is worth reading even when you think the conclusion is wrong, and one that hides them is worth nothing whatever it recommends.

Table diagram of reading a bank research note: skip the rating and price objective, read the assumptions.
A diagram of this page's advice: the assumptions are testable and the conclusion is not.

What an investment bank is not

It is not a high street bank. No deposits, no accounts.

It is not your adviser. The client is whoever pays.

Its research is not a forecast. It is a published view.

And a price target is not a plan. It has no invalidation.

How this connects

Goldman Sachs is the name most people picture. The largest firms combine advisory work, underwriting and trading desks, while smaller boutiques do only the advising.

Underwriting covers far more than common stock. A corporate bond or a convertible bond issue runs through the same process, and companies come to the bond market to borrow far more often than they sell new shares.

The buyers are mostly institutions. Asset managers such as BlackRock take large allocations of a new issue, which is part of why a retail investor sees the price only once it trades.

The research leans on the same tools a private investor uses. A DCF model and a close reading of the 10-K sit underneath most valuation work, and an analyst’s questions on an earnings call are often more revealing than the rating.

Trading desks sit close to other kinds of firm. Several large banks operate their own dark pool, and the speed-driven side of market making overlaps with high frequency trading firms.

The largest banks are a source of systemic risk. They are counterparties to one another on huge derivative books, so counterparty risk at one can travel through the rest.

And none of this makes a bank a fiduciary. A financial adviser acting under that standard owes a duty to you, while a bank’s duty on a deal runs to the client paying for it.

When it fails

In a quiet market the fee income disappears rather than shrinking. Deals do not happen slowly; they do not happen, and a business built on transactions has revenue that goes to nearly nothing.

The second failure, for a reader, is treating research as advice. It is written for institutions with different horizons, different sizes and different constraints.

A third is buying a flotation because the bank priced it. The pricing served the seller.

A fourth is following a target without an invalidation. There is nowhere to be wrong.

A fifth is assuming the trading desk shares the research view. They are separated deliberately.

And a sixth is inferring anything from a rating change alone. The information moved before the notice did, which is what the gap on the chart already told you.

The original data

Of the 24,971 videos in this site’s search study, 2 have “investment bank” in the title, at a median of 793,650 views from a single channel, with a maximum of 1,125,920. Hedge funds appear in 22 at a median of 83,975. Duplicates are removed first, so no video is counted twice.

The question: A buy rating from the firm that sold the deal. Weigh it?

Two videos from one channel, and a median of 793,650 views apiece. That is among the highest medians measured on this site from among the smallest supplies — the institutional side of markets is watched avidly and explained rarely, because there is nothing for the viewer to do with it afterwards.

The answer to that final question is to weigh the note and discard the rating. Read what the analyst assumed and test those assumptions yourself; the rating was produced by a firm whose fee has already been earned. If the assumptions are not stated, there is nothing in the note to weigh.

Hedge fund is the counterparty on much of this activity, and the other private side of the market. Market makers is the trading function described on its own terms. And stock exchange is where the issues these firms underwrite end up trading.

What I actually do

I ask who is paying before I read anything a bank publishes. When a bank prices a flotation, the company selling the shares is the client, and getting a high price for them is the job done well. Nobody has to be dishonest for that price to tell you nothing about value.

— Michael Whitman

This page is educational, not financial advice. Test every idea on your own charts before risking money.