By necessity, many non-bulge bracket and most other investment banks have turned to private placements through their broker-dealers to make up for the lack of fees that typically have come from IPO work. A trend has emerged for the investment banks to apply public company underwriting parameters to private company placements, as the banks enter into the largely unfamiliar (to them) area of venture capital financing as agents.
The problem arises when the banks migrate provisions from agreements used in, for example, their public registration activities to a typical private placement deal. Perhaps the most provocative example is anti-dilution protection in a warrant in the event of a subsequent "down round."
In pure economic terms, it is not at all clear why a pure placement agent (compared to principal investor) in a Series A round should get down round protection at all. The agent is not taking any significant risk (the financing is not underwritten) and not putting up any money. However, not only do some agents insist on typical protection, it is sometimes full ratchet and, adding insult to injury, constitutes a "double dip"... not only does the warrant holder get more stock (as does the convertible preferred), it pays less money.
While the fee arrangement between the placement agent and the issuer is negotiated between themselves, it is a cost that is ultimately borne by the investor. Therefore, when a placement agent is used, it would be useful for investors to know not only how the agent will be compensated, but perhaps more important, what role the agent plays. There is a wide variance of sophistication level among investment banks acting as placement agents. Some act simply as introducing brokers. Others are more value-added advisors, counseling management, revising business models, etc. In the latter case, it certainly makes sense for the agent to participate along the same lines as other investors. In the former case, where the agent is a true introductory intermediary (and nothing else)--the decision depends on how badly capital is needed.
Needless to say, the details of the terms can have significant effects on a cap table. Repeating the Note on Full Ratchet and Weighted Average in the VC Encyclopedia, full ratchet is the provision wherein the exercise price goes down dollar-for-dollar... a $5.00 exercise price goes to $4.00 if the down round is at $4.00.
If weighted average is the formula, then the new exercise price equals the old exercise price multiplied by a percentage (less than 100%) which divides the sum of pre-placement shares outstanding (say, 100,000) and the shares which would have been issued had the old, i.e., higher, price been paid by the sum of the pre-placement shares outstanding plus the shares actually issued. The shares actually issued being a bigger number than the shares which would have been issued, the percentage is less than 100% and drags, therefore, the exercise price down... but not way down because of the ameliorative effect of using the outstanding shares. To illustrate, if 12,500 shares are issued in the down round (at 4.00/share) and $10,000 would have been issued had $5.00 been the price.
But, then, with the holder paying less money, she also gets more shares. The usual formula is the number of shares subject to the warrant times a percentage that is greater than one. For example:
# of shares Subject to the Warrant = 100
Old Exercise Price = $5.00
New Exercise Price = $4.00
$5.00 / $4.00 = 125% X 100 shares = 125 shares
Hence, for less money, the holder should get 25% more shares.... 125 instead of 100.