I’ve gotten a lot of responses to my last column (it’s right below this one if you haven’t read it). In it, I laid out 5 principles for any bond market system, and briefly analyzed a few potential solutions. With the rash of audience participation (both by email and on TT), I decided to examine the multitude of proposals in a more formal setting.

These were the 5 principles (explanation is offered in the previous column):

Principle 1: There should be adjusts based on box office performance.
Principle 2: Adjust prices should not depend on the current price of the bond.
Principle 3: Bond ratings and prices should be able to go down if a movie bombs.
Principle 4: There should be a minimum of HSX judgment involved.
Principle 5: It should be understandable.
And one that I forgot:
Principle 0: If A > B, then the adjust price for a given bond if the movie grosses $A should not be less than the adjust price if the movie grosses $B.

In order of importance, it should go 1, 4, 0, 5, 2, 3.

Richard M. writes:

Actually, you seem to have an additional principle implied: Principle 6: A bond should adjust after every movie (or you could say that this is automatically true of any system, in which case this is an extension of 2). For the record, I don’t agree with principle 6. I don’t object to systems that meet it, but I don’t think it’s required. I would restate it as: Rich’s Principle 6: It should be possible for a bond to adjust from a movie, if the movie grosses a certain amount. Obviously, this is not true of the current system either.

He’s correct – principle 2, taken to its logical end, bans the “No Change” adjustment. Rich’s principle 6 is a milder version of 2. Rich goes on to take issue with number 3:

I don’t really agree with this principle. After all, what does a movie have to make to be a bomb? Less than a million? Less than 5 million? Does it depend on the movie? Further, this principle could result in situations in which a star’s TAG goes up and the price adjusts down. For example, under some system that satisfies this principle a star who has a TAG of 0.2 is in a movie that grosses 0.5. His TAG rises to 0.35 and his price drops to $500 (this is a hypothetical system, so I’m making numbers up). The star has another movie that grosses 0.5 also, which brings the star’s TAG to 0.4. Since the movie bombed again, the price would have to drop again, to say $400. It’s also worth noting that this principle is mutually exclusive with principle 1 (after all, if a bond has to adjust to a fixed price, then it is impossible to guarantee that it always drops in price after a bomb).

I think I bungled the explanation of what exactly I meant by principle 3. What I meant to say was: after an adjust to $X, there should still be a possibility that the next adjust will make the bond adjust to $X or less (and vice versa, I suppose). Hmm. Now that I say it correctly, it seems even less important than it initially seemed. Oh, well.  

The Solutions

Keep the current system.

Beast writes:

With all the talk about a new bond system, no one has really explained what is so wrong with the current system that necessitates a new system. It seems to me (remember that I’m not a bond guru) that the current system is works just fine. With the advent of V2, traders can profit from upward and downward adjustments–bringing that much more activity to the bond system. Also, with stop orders, traders don’t have to sit at their computers on Thursday night waiting for adjusts to happen. They just place the stop orders and that’s it. It seems as though the main problems associated with the V1 bond system have been addressed with the V2 bond system. However, from what I gather, the main argument currently being made against the current system is that a bond can adjust down if it is attached to a blockbuster movie (e.g. EMCGR in STRWR). My counterargument would be that the market has simply overprice EMCGR, and that the adjustment is merely correcting the inflated price. So, what other problems are there with the current bond system that necessitate a new system?

I agree with everything above. The only problems currently that I see are that principles 2 and 3 (and 0, sort of) are violated (with the exception of timing problems – when to schedule the adjusts, whether to halt bonds, etc. – which I’m ignoring for the purposes of this article). Maybe if Mac (are you reading this, Mac?) could enlighten us as to why exactly he feels the bond market needs changes… I’m not opposed to changes, though, as long as they’re done right.

Current system, except “No Change” adjusts are replaced by adjusts to $1000.

No one raised any objections to this one.

Prices for each rating. If a bond is U after a movie, adjust it to $500; if C, $600; etc.

Dave Silberman sez:

Problem with setting a price for each rating: a bond that is AAA and is likely to stay AAA (THANK, GLUCA, for instance), will be a horrible investment, because it will never move. If it moves up, it’ll be time to short, since the next movie will bring it *down* to AAA.

True enough. Although the same is true of the current system, is it not? Right now, any AAA bond over $800 is a bad investment if you are thinking just about adjusts. No ratings; just adjust bonds to $X + $Y * TAG.

I don’t like this one any more. If an actor is in a lot of movies, that means that the swing in TAG isn’t going to depend much on his next one. Even if people can’t decide whether it’s going to delist at $10 or $100, if he has 9 movies previously, then the predicted TAGs aren’t going to differ by more than $9. This means that the range of possibilities for an adjust value won’t be large enough to worry about, especially since the range will be narrowed considerably after the first evening’s numbers come out. The only remedy for this is to make Y very large, like $50 or so, but if it’s that large, then we’re going to have bond prices in the $8000+ range. Yikes!

Some other suggestions:

Preston C. wrote, “[C]hange all the prices to the TAG of the star.” This is essentially the same proposal as above.

Mara on TT suggested that bonds adjust to the sum of the box office of their last 5 movies. I think this suffers some of the same problems as the last proposal, but with some modifications (use the sum as the TAG, sort of like director bonds) might work. As there aren’t too many problems with the way TAG is computed, though, I’m not sure this would be an improvement in and of itself.

OsbOrne wrote to suggest the following:

1) A rating system for the actors in the film (the headliners would get top billing, extras the lowest billing) how many positions there are in the rating chart I couldn’t hazard to guess… 2) Using the above ratings, assign a percentage of BO takings to actor.. (top billing say nominally accounts for 25% of BO gross) 3) Bond value calculated based on for arguments sake last 6months BO takings (rolling on a weekly basis) using the above rating system multiplied by some figure to give a sensible figure for a bond price…

The problem with this is that a bond automatically goes up when a movie with that actor opens, and it goes down 6 months after said movie. Also, it would be a lot of work to adjust every bond every time, and I don’t know that HSX is up to doing number 2 (they have a tough enough time just figuring out who’s in the movie). Redbird posted the following on TT:

1) The first priority would go to the normal system as it now exists. If a bond adjusts to H$714, it’s done. 2) But for the TAG, if it doesn’t reach the base level, it would go down H$1000 and if it goes above _without adjusting TAG-side_, it would go up H$1000. So for XYXYX, a BBB bond that would not adjust under Priority 1…if it misses adjust but does less for this movie than its TAG (in this case $39.9 and less), he would go down a thou points. If it makes $50 or more, it adjusts up a thou.

The problem I see here is that principle 0 is often violated. Can someone think of a good workaround? Also, what happens if neither 1) nor 2) is satisfied? I haven’t seen a perfect proposal yet. Hopefully, the ideas that I’ve written about here will inspire more. Right now, I think I’m partial to giving each rating a price, as I think it’s simple and doesn’t run into any paradoxical situations (as far as I can tell).

It can be blended with Redbird’s suggestion, if desired. For example, say the rating prices are as follows:

U : $500
C : $650
CC : $825
CCC : $1000
B: $1200
BB: $1450
BBB: $1700
A: $2000
AA: $2500
AAA: $3000
Say we have a B bond. If, after a movie delists, the TAG dips to CCC level, then it adjusts to $1000. If the delist * 1.24 is within the B range ($20-$30), then the bond adjusts to $1200. If, however, the delist * 1.24 is less than $20 but not low enough for a rating change, then the bond adjusts to the midpoint between B and CCC ($1100). Similarly if the delist * 1.24 is higher than $30 but not high enough for an upward rating change, it adjusts to $1325.

As always, criticisms and suggestions are welcome.