2008 Jul 12 - Sat

Freddie Mac and Fannie Mae

The markets are not very patiently waiting for all the sub-prime mortgages to reset later this year and next year. I've always thought that it will be during that time period in which we'll find out if we are in a repression, depression, or a recession.

It seems we have lots of near term worries. Various and sundry resets must already be taking place. The financial system seems to be stressing all over the place. I'm curious to know just how leveraged the industry is. After all, in the end, the money has to be some place. If leverage is the problem, someone is going to be left holding the bag, and I guess that is starting to happen now.

This reminds me of the great internet expansion a number of years ago. Every who thought they knew what a fibre optic cable was, starting laying the stuff across land and under water. Most if not all of those companies reorgainized or got sold at pennies on the dollar. The smart guys on the sidelines smiled and spent their pennies on valuable infrastructure. Mean while the original investors and vendors probably didn't fair so well. in the end, after the market settled, we have high capacity bandwidth at reasonable prices (well for North America anyway, we in Bermuda still pay an arm and a leg for the privilege, although that should change with the new consortium laying new fiber later this year).

That story leads me into today's newsletter by John Mauldin. He is saying that sub-prime resets aren't our only problem. Orbourous is eating its tail. Lenders to lenders and lenders to corporate beings are starting to cause problems. A figure like $1600 billion dollars of losses in the international banking system are being bandied about. That is a terrific amount of business failure, residential collapse, and bad business judgement. Is there a figure somewhere that suggests what the equivalent of a world wide Gross Domestic Product might be? (I know that appears to be a contradiction of terms, but with the ever shrinking world, there is an element of realism there). With a WW GDP, this type of loss could be put in to perspective. Grasping at straws, I came up with one form of perspecitive. According to the US Federal Reserve Statistical Release, that is $300 Billion more than the M1 money supply and about 23% of the M2 money stock measure.

Anyway, his previous articles had some concrete examples as to what sort of numbers losses were based upon. In this article, there seems to be a bunch more hand waving going on. Perhaps the Bridgewater Associates report to which he refers offers up some concrete basis for their opinion.

Finding the headwaters of investment sources is what John Mauldin's friend David Kotok specializes in. In a recent newsletter, he is saying that Freddie and Fannie (F&F), between them, hold about $5000 Billion in mortgages. Me, coming from the outback, think that a $150,000 mortgage is big. Having one of that size, and if I've used the correct number of zeros in my calculations, that could mean about 30 million mortgages. To stretch the statistic even further, that would be a mortgage for 1 out of every 10 US residents. That is a lot of cash flow to them and to their holders of sub-paper.

Both of the authors tend to agree that holding paper from F&F is not too risky being that the mortgages that they do hold are reasonably sturdy, and they both agree that holding shares is valueless. So, so long as the cashflow meets payment expectations, things shouldn't be too bad.

However, all this is contrary to what the notable publications such as WSJ are publishing, so no wonder we bounced off the Dow 11,000 level yesterday. I think Mauldin even joked about the 9XXX level not being too far off the mark in his article.

One other thing Mauldin mentioned is that he is doing a survey. As a reward for filling out his survey, he provides a link to speech in which he talks about how the markets might re-arrange themselves. Perhaps this might be similar to what happened with the post fibre-laying companies... will the new credit/debt institutions be valuable because of what they got for pennies on the dollar?

I wrote this article in order to set a baseline of expectations of what is to come. Will we, indeed be seeing more losses, more than what the subprime fiasco has caused directly? In which direction are the markets headed and what will be their prime motivator? Will it be more credit problems? I'll be able to look back here and hopefully see what happened when we start our descent into the 10K category.

[/Trading/MarketNotes] permanent link

Open Source: Sonic Visualizer

This blog entry should actually go under the title of Open Source Package of the Day, but because it solves a different problem for me, it comes under the heading of Lighting.

Back in April, I did lighting for a student Jazz group. As part of the show, I wanted to do my own little ditty. My desire was to do a mini light show set to music. In some shows, I've seen the lighting designer loosely sync the lighting display to the music theme. My goal for this particular display was to ave the light show visibly keep time to the beat of the music.

I considered quite a number of ways to do this. Some more manual than others. A key feature had to do with my inherent 'laziness'. Why should I have to manually redo and retranscribe the beat when I should be able to get that out of the music file itself? Easier said than done. I first turned to Cakewalk's Sonar Producer Edition to do waveform analysis for me. With the waveform editor, it is easy to find 'note starts'. Those note starts don't necessarily carry the rhythm. Another tool within Sonar allows one to filter to certain frequencies. I was interested in the bass beats. That helped, but would still very time consuming to identify and place a note by hand.

I finally cheated. I found a fully functional MIDI file of Pinky and the Brain. All the instruments were nicely laid out, plus it had a kick drum track. This kick drum track was monotone, as a kick drum track should be. This made it dead easy to set off cues from the MIDI input in Light Factory. I ended up with a 20 to 30 second subsection with 157 cues.

The little ditty worked very well and the audience loved it.

Now that I've figured out music timing, light timing, and integrating the two, not to mention that amount of time one can spend on making it look easy, I've been thinking of turning the concept up a notch, and maybe doing a larger production.

Up till now, I figured I'd only be able to this with MIDI tracks, as the timing and instruments are all laid out, and I really didn't/don't want to go back to waveform analysis, at least being limited to what Sonar provides.

Enter Sonic Visualizer. With this, one can take a mucic track and run the standard note start tools on it. The cool feature is the spectrogram views. One can actually see the vibrato of the violin, the beat of the drum, the complex weave of tones of a symphony, .... The program also has a MIMDI annotation feature.

I'll have to manually key in the MIDI notes I want for cue changes, but with the visible segmentation of notes, instruments, and voices, creating and coordinating a lighting show tied to music could be a delight to do. Lots of time will be eaten, but with Sonic Visualizer, Light Factory, and a lighting show visualizer like Capture Console, it could be quite an experience, not only from a design point of view but from an audience pleasure perspective.

Ok, back to daydreaming. In a previous article, I wrote about a light console. Not any ordinary console, but one which is simply a piece of acrylic with infrared sensors on it. This would allow realtime multiple touch live input. By including this in the show, on stage, a light show would become live performance art in itself. I think a live spectogram would be an interesting light show addition in itself. Are there such shows out there?

[/Personal/Lighting] permanent link