mini-review: AudioEngine W1

I twittered about the AudioEngine W1, and having had a couple days to try it out, I thought I'd post some notes.

Basics first: uncompressed PCM audio from your PC to stereo, wireless. My goal was to connect an electronic keyboard to a stereo system across the room, which isn't really their intended application, but was worth a shot!

Pros:

• Setup is great: USB plug and play (with XP at least) required no drivers
• They have a "pairing" procedure if you add more hardware to a system, but when you buy a sender/receiver they come pre-paired.
• No pops or clicks when connecting/disconnecting
• Makes your laptop feel like magic
  
• Sound quality is quite good (I haven't done a detailed quality test yet, but my ears say so)

Cons:

• Clipping! From my Roland RD-700 to my stereo, I've had to leave the volume at 70% to avoid clipping. This might be a big problem if you have a device with line-outs and no volume control.
• Latency: for musical instrument use, the latency of their system is noticeable. It's not awful, but likely 100-200ms. Enough to mess you up if you're trying to play.
• Only one USB power jack in the box (I scavenged one from an old phone), so they expect you to be using a USB-capable device as either sender or receiver: e.g., sending from a PC or receiving with their powered speakers. They really should ship two, because these things are pretty hard to find.
  
• Cords are unshielded, and they ship only one RCA adapter, and two stereo 1/8" adapters. I already had all the cabling I needed, but still, another RCA would have been nice.
  

Overall, I could recommend this strongly for their intended application: wiring music from your PC to a remote speaker/stereo. For use as a remote instrument, not so much.

Compared to the AppleTV with Airtunes, the AppleTV has "more" bass and louder output, but otherwise the DACs seem to be of similar quality. If your only app is iTunes, it's hard to beat Apple's solutions. If you do other things with your computer, you'll like this device.

Still, for now it's pretty unique: at $150 it does beat the rack-mount $1000 competitors. It's stereo and isn't powered by batteries like guitar amps. It's digital (not X10/awful) and the quality seems pretty decent. So it seems to me that the AudioEngine W1 is breaking new ground at this price point, and it's quite a nice start.

Printing Bad Money

Michael Lewis (of Liar's Poker fame) wrote about the housing crash: The End of Wall Street's Boom.

By far the most amazing point in his exposition is how he describes how CDS was used to create leverage. Seriously, you won't believe this.

Here Lewis is describing Steve Eisman (who was shorting subprime BBB tranches using CDS):

Whatever rising anger Eisman felt was offset by the man’s genial disposition. Not only did he not mind that Eisman took a dim view of his C.D.O.’s; he saw it as a basis for friendship. “Then he said something that blew my mind,” Eisman tells me. “He says, ‘I love guys like you who short my market. Without you, I don’t have anything to buy.’ ”

That’s when Eisman finally got it. Here he’d been making these side bets with Goldman Sachs and Deutsche Bank on the fate of the BBB tranche without fully understanding why those firms were so eager to make the bets. Now he saw. There weren’t enough Americans with shitty credit taking out loans to satisfy investors’ appetite for the end product. The firms used Eisman’s bet to synthesize more of them. Here, then, was the difference between fantasy finance and fantasy football: When a fantasy player drafts Peyton Manning, he doesn’t create a second Peyton Manning to inflate the league’s stats. But when Eisman bought a credit-default swap, he enabled Deutsche Bank to create another bond identical in every respect but one to the original. The only difference was that there was no actual homebuyer or borrower. The only assets backing the bonds were the side bets Eisman and others made with firms like Goldman Sachs. Eisman, in effect, was paying to Goldman the interest on a subprime mortgage. In fact, there was no mortgage at all. “They weren’t satisfied getting lots of unqualified borrowers to borrow money to buy a house they couldn’t afford,” Eisman says. “They were creating them out of whole cloth. One hundred times over! That’s why the losses are so much greater than the loans. But that’s when I realized they needed us to keep the machine running. I was like, This is allowed?”

I keep reading that again and again.

This is how bad it got.

They actually "cloned" bad loans (by some estimates 10:1) because the "returns" were so great and people wanted to buy.

Monstrous.

Vista Uncapable Intel

Lots of chatter around the emails published due to the "Vista Capable" lawsuit. It's of course upsetting that Intel could push their 2001-era technology in 2006, and reading the emails, it's amazing to me that they didn't know their hardware just...sucked. Overall, I think there should be a lot less criticism of Microsoft in this and more scrutiny of Intel. Having your CEO ask for favors that deceive customers is very bad indeed.

For years people would ask, "Why don't you use graphics hardware in Picasa?" and we'd always say "because it doesn't work on the lowest-end PCs" and largely that meant Intel's awful integrated chips.

I suppose this is what you get for squeezing margins on PCs too much...Apple didn't ship Intel's integrated chips, ever, and instead used discrete chipsets in all the models from 2002 on. That choice probably cost them an extra $10 per machine, for which their users got 10x the graphics performance.

But in a backwards-looking way, I can criticize Microsoft for not providing a software path that implemented Aero Glass in a reasonable way. In 2006, a new Intel CPU could do a pretty good job of full-screen compositing. Maybe there were some doubts about this as Longhorn was ramping up earlier in the decade, but a dual-core 2GHz machine in 2006 was capable of amazing, amazing fill rate, seriously.

Apple managed to pull off a full CPU-based path in 2001 using just a 400GHz G4, and while the performance was pretty bad on a slow CPU like that, Vista's performance would have been quite good in 2006, and it would have worked on the old hardware that was being shipped then. And it would run on ultraportable machines with tiny screens, bad GPUs, and good CPUs.

But the silver lining here is very good. By kicking the awful integrated chipsets out of bed, by 2008 the spread of performance from the "fastest" to the "slowest" hardware in the industry has narrowed, perhaps from 100:1 in 2005 to 10:1 in 2008.

This means that on average, if you're making a new graphics-intensive app in 2008, you can rely on decent hardware support. And this result is almost solely due to the Vista requirements. Awkward process, but it has been extremely good for software developers.

Electric Energy Costs

I noticed our electric bill was a bit higher than usual so I ran the numbers against a bill from two years ago.

For LA DWP, costs are up 15% in the last two years, from 10.4c/KWH to 12c/KWH. With taxes included, that's 13.2c/KWH.

Maybe someone finds that useful...

Savings Insurance?

I've been thinking about the really heart-wrenching stories in the housing bust.

The people with 0% down, who can't afford the house, have a non-recourse foreclosure, and then go back to renting don't tug my heartstrings. They were effectively renting, anyway, and I'd prefer to see our government staying away from the landlord business, with a goal of creating real value and ownership for people in their homes.

The stories that seem more awful are people who used their house as their primary means of savings. A retiring couple that put a huge amount of their net worth into a house, either by purchase or improvement. And I don't mean home-equity loans used for improvement. I mean, actual savings, spent in a declining market.

I'm contemplating what it would mean to have the government provide "savings insurance" that would guarantee, say, 70% of the equity in a home when it was transferred from some liquid vehicle. 20 years of mortgage payments would qualify, as would a 30% downpayment that was lost in a declining market. Cash-out refi's? Excluded.

I've not heard this proposal anywhere else, so I'm just writing it up to see if it holds water.

Maybe someone can make a counterargument, because I think this plan seems pretty reasonable. Though I can't estimate the cost, it rewards the right things. If you don't correct for inflation, it would have the effect of just "fixing" people who lost significant money due to the housing bubble, instead of people who bought in the last one, for instance.

Sheila Bair's Bailout

Sheila Bair runs the FDIC, and she was tasked with handling IndyMac's 60,000 foreclosures (out of approximately 700,000 total loans). In an effort to reduce losses, she's proposed some aggressive workouts for homeowners in foreclosure.

Things like: preserving teaser rates for another five years, or reducing loan balances to 90 percent of "market value". These terms are actually similar to Countrywide's California bailout (which was motivated by a lawsuit, not loss mitigation).

But now, the IndyMac plan is being held up as a model for the entire country (LA Land link here, with interesting comments too), and that's where it gets dangerous.

I've debated the pros and cons of this proposal with a few friends, and I have a few thoughts:

1. The model seems to be that most markets have experienced their losses, and now we'll have a V-shaped or shallow-U-shaped recovery. We made bad loans, and now we don't, and so we can pay those off and be done with it. But this is not true. Houses are overpriced in some areas by almost double, and their prices need to fall more.

2. What do future losses mean? If these losses are still to be taken, it is dangerous to prop them up by reducing foreclosures. Not because it "reduces losses" in the short term (that's nice if you can have it), but because new underwater loans will be written on the 2008, 2009, 2010 homes, and will remain underwater for 10-15 years. This will make lending hard for everyone else, and will mess up the credit markets for years to come.

3. If it's extremely easy for a 60-day late borrower to adjust to half the previous payment, we'll see a huge rise in pre-foreclosure activity, with borderline borrowers failing to make payments so they can "qualify" for the bailout. (Peter Schiff advised people to do this on CNBC!) Costs for Bair's plan will double, triple, quadruple...we'll have half the 2004-2007 loans qualifying for 2% loans. There will be two classes of homeowners: those paying rent to the government, and those who use real money to buy and sell.

4. It's also nice, if you can do it, to avoid massive foreclosures, because typically foreclosures will "undershoot" correct pricing and cause destruction of assets, due to lower selling prices and non-maintenance. However, "normal" 5% down homeowners cannot afford to take 50% losses on property, and buyers cannot afford to pay more. If we read our history from Japan and California 1990-1995, sellers mostly think in nominal prices, not real prices, and so they will take an inflation-sized "real" decline, or maybe sacrifice a small downpayment, but not much more, not bankruptcy. Asset prices do not fall when they're propped up, they decline a few points beyond inflation for years and years. There are plenty of buyers at the right prices, and properly-priced homes will simply not be abandoned.

If we want every housing investment made in the next 5 years to be immediately underwater, but minimize short-term losses, with a ridiculous split between what sellers can afford to sell for and what buyers can afford, Sheila's plan is good. But I think everyone should agree that this is more short-term thinking, not long-term investing in our financial health. Even more borrowing against the future...the last gasp.

Ten years ago, 64% of the population "owned houses". Last year that number was 74%. I do think that some of the 10% of the population that have become homeowners in the last 8 years will probably need to become renters (not propped-up homeowners), and that a 2% teaser rate that resets in 5 years is not doing anyone favors. Pushing off a new round of teaser rate resets to a time in the future seems wishful...the only way this plan works is if we have enough inflation to burn off the excess in 5 years or so...imagine 1980's-scale inflation (15%).

Wells Fargo is selling its foreclosures into the market immediately, not trying to "spread out" losses over some future time. Why? Because it's actually smart if you can sell at elevated prices today and avoid selling at reduced prices tomorrow.

I would rather see us take our lumps now (these numbers are known and manageable), and have a healthy system for the future, rather than pretend that current asset prices will come back, and save a small portion of our losses this year, at a cost of messing up the system for the next ten.

Juan Enriquez on the Economy (PopTech!)

Juan Enriquez (2008) Pop!Tech Pop!Cast from PopTech on Vimeo.

Electoral politics makes Bad Economics

Interesting commentary from Peter Viles at LA Land:

Yet the state's economy and specific problems go undiscussed in presidential politics because its votes are already counted for the Democratic Party. If Americans really knew about the details of our wacky housing market -- median home prices above $500,000, house painters buying brand-new $375,000 homes with no money down, etc. -- they would probably be shocked.

The ideas voiced by McCain and Sheila Bair about propping up home prices sound totally insane to Californians -- we have a real bubble and prices need to go down. But in an election year, California doesn't matter because the votes are already counted.

Policy seems to shift just enough to be what people want to hear in the swing states, and not what the dollars say about 34% of stupid mortgages made (12% of the population, 34% of the mortgage dollars in California) in the last 5 years.

Central-limit Gaussians

This is probably more of a stereopsis.com post, but given that this is short idea, I thought I'd just put it here.

iTunes plays all my old music with the same priority as new music, so I wanted to talk about a really simple way to make random numbers that favor recent things...because it's easy! (We did this in a few places in Picasa.)

Mathematically, the "central-limit theorem" says that if you add together a bunch uniform-random samples, you get something that looks like a Gaussian, which favors stuff around its middle, while still providing some likelihood that "tail" values will get picked.

So I want iTunes to do this:

if randx() returns uniform-random numbers in the range -1, 1:

gaussian = fabs(randx() + randx() + randx()) / 3;
iTunes.nextshuffle = dateadded[gaussian * numsongs];

Maggie Votes

Maggie and Kanye

We got Maggie a cool Halloween costume, so she can look like Kanye.

PRI's The World on lending in the middle east

My favorite part happens around 1:35 or so.




The Middle East has so far seemed to escape the financial crisis plaguing the rest of the world. As The World's Aaron Schachter reports, this is in large part due to inherent wariness and a banking and real estate boom of its own.

The Great Depression Index

Google Trends gives tells us this about people searching for "Great Depression".
Leading depressed cities: Irvine, Chicago, Philly, LA, Phoenix, Houston. Wait, that's where all my friends live.

Predicting the Future

I've been reading blogs from economists and finance people and policy people, and I've had a lot of time just to think about things. It turns out that when you have a lot of time to think, you notice how your own head works and think about that, and so it goes, in circles mostly.

I went to a figure drawing class tonight, and the mindset there is "Now." As in "Now, I'll draw this curve," or "Now, I am noticing how the shape of this body part works with the shape of that body part." Not too much anxiety about the future or the past. Well, aside from the "I'm doing it all wrong," frustration, but otherwise not really.

This post is about "Predicting the Future" because that's the mindset I've been in, and that's the mindset of these economist guys. And I've been convinced that the economy's going to be bad any minute now for the last two years, and so now I'm "right" and it's not much of a victory.

Before that, I was Managing a Software Team and thinking about what could go wrong and what was risky and what was easy and who was going on vacation and would things get done on time, and we need more resources over here because otherwise we can't do both of these things at once. It's a more definite form of Predicting the Future, because at least I'm an expert at making software, whereas with economics and finance I just repeat what I read other places, and I don't practically do anything about it, except not buy houses and sometimes buy equities.

But it's not like Making Stuff, which is kind of like Now, and kind of like Seeing the Future (as opposed to Worrying about it).

And the thing I've noticed is that Predicting the Future (aka Worrying About the Future) gets in the way of actually Making Stuff.

If you have a company and you have meetings mostly about what can't be done, or you feel like you're working around constraints, you're working at 20% of your real speed. It doesn't matter how awful your company is or how many constraints there are, but you could get 500% as much done if you'd just do it, instead of worrying about the future.

Note I didn't say never worry about the downside, I said don't worry about it at the beginning. A few months ago, I was talking to a good friend whose company wasn't doing so well. He said, well we have this process, and it seems very organized and it makes sense...but his company wasn't doing so well. I just asked, "What's the downside if you fail?" Not much. "What's the upside if you do something great?" A lot.

Some companies and people have a lot to protect. And if the goal is to Maintain the Status Quo (one might say that this is mostly Ben Bernanke's job), there are a lot of things to worry about and a lot of ways to worry.

But if you're a person for whom Seeing the Future and Making Stuff are more important than knowing all the things that can go wrong all the time, then you should take some time every so often to turn off the worrying part of your brain and do things that are creative and worry-free.

I've known some people who manage to dance on both ends of this see-saw effectively. Really smart lawyers, software security geniuses (clever & careful at the same time). But mostly people fall on one side or the other. I don't think the human brain can fully do both at once.

But I know I can't do it. I can't Predict the Future and do my best work all at the same time. When I'm Predicting the Future, I say, well, what if this doesn't work on all platforms? And I don't start, and I never reach Now.

I believe this is why a lot of successful startups often are run by a pair of founders, where someone can do the worrying and someone can do the non-worrying. But I probably need to kick myself back into the old non-worrying camp and see what happens for a while.

Countrywide (BofA) refinance details

Countrywide and BofA have provided some details on their refinancing plans for subprime borrowers.

It appears these modifications are possible for buyers who owe at least 75% of the value of their home, and who have missed 60 days of payments and who bought sometime in 2005-2007. And why not "miss" 60 days of payments? Look what you can win!

Terms mentioned include:

1. All borrowers are encouraged to switch to an FHA loan, but why would they? because...
   
2. Option-ARM borrowers may have the principal reset to 95% of the current market value of the house and have payments reduced to 3.5% rates.
3. ARM borrowers may elect to keep the "teaser" rate for 10 years, or use an interest-only loan at 3.5%.

This rate is about 1/2 the payment a full-amortizing 30 year fixed loan would require.

If sellers are paying 50% the payment that buyers are paying for the same house, then it will take approximately 15 years for inflation (at 4-5%) to make up the difference in pricing expectations.

Apparently Countrywide was servicing about 400,000 loans from this period, and they figure they'll pay about $8,750 per loan. ("Fixing" a 40% price decline for the same loans would cost them about $88,000 per loan instead.)

So their numbers sound really low to me, but I actually fill out rebate forms, so what do I know? Maybe they'll make the paperwork hard this time.

Mortgage Equity faucet turned off: is there a college price bubble?

Calculated Risk updates with the latest Kennedy-Greenspan numbers for Home Equity Withdrawal (aka Home Equity Loans).

From a high in 2005-2006 of nearly 10% of disposal household income (think about that for a second), we now see 0.3% in the last quarter. This is a massive change in domestic consumer spending. In my view, unemployment isn't the only indicator one should look at to understand recession.

A second thing that's been rattling around my head for a while: I'm wondering actually if college costs (which have been greatly outpacing inflation) will run into the same housing-price wall? I expect many parents sending kids to college in the last few years have paid for it via Mortgage Equity Withdrawal. Someone with real data should run those numbers.

Don't....vote.

FDIC move to insure $250k passes the senate

The Senate just passed (as part of the bailout bill) a provision to raise FDIC protections from $100,000 to $250,000 per account. This merely restores the inflation-adjusted protections from 1980, as I noted in July.

Wired makes the accusation that this "insures deposits for the wealthiest" but there's more to consider. According to HousingPanic, fully 1/3 of bank deposits were uninsured by the old limits.

This change avoids a run on banks for a considerable portion of the $2T that was previously uninsured, hopefully at a small cost. The "bailout for the wealthy" story needs to take into account at least some of the systemic risks, but that's been the story all week.

Housing prices across cities: 10-year data

I've been looking for a good metric to compare prices across cities, to estimate where the housing boom "happened" the most. It seems like the data becomes cleanest at about 10 years ago. For instance, in comparing cities in California: in 1998 Los Angeles was coming out of the early-1990s housing price crash, whereas San Francisco hadn't realized many of the rewards from the tech boom.

If you wait until 2000 (which Case-Shiller is based on), LA and SF have diverged wildly (which makes SF look less bubbly than LA). LA's growth doesn't start until after 2000, whereas Silicon Valley has a pop in 1999 and 2000.

The graph that convinced me of this is First Republic Bank's "Prestige Housing Index": Here's LA and SF. Quite a difference. (FWIW: I think this index is pretty much bunk back if you go back to the 80s, because they thresholded properties costing >$1 million, and there just weren't that many of them back then.) Since the 90's, though, things look interesting, and the "prestige" data bears out my claim that Prime areas haven't seen their doomsday yet.

Whether this means that Silicon Valley/SF will retain some of their gains, whereas LA will lose more of them, I don't know.

But, this is one of the reasons I like the ten-year Zillow graph to compare zipcodes. A lot of the stock market madness from 1999-2000 seems to be canceled out if you go back to 1998.

Bottom-up design....and subprime

I didn't really expect to be relating software to subprime "affordable housing".

But I read Mike Tsao's Angry Rant about the Mortgage Crisis and find myself inspired.

The American Dream is living in a country where you get to go to school instead of working in a factory through your childhood. It's having someone answer the phone when you dial 911. It's getting a job and keeping it in spite of the fact that your religion, race, place of birth, or last name is different from that of your boss. It's having an impartial court instead of a baseball bat to enforce legal contracts. It's starting a business without having to bribe corrupt officials.

The American Dream is not about having. It's not even about earning. It's about living in a place where earning is even possible. It's certainly not about promising to pay a mortgage of $4,000 a month when you earn $55,000 a year because the "american dream" entitles you to do so.

He's spot on, but more than that, affordable housing should be attacked by the old proverb: “Give a man a fish; you have fed him for today. Teach a man to fish; and you have fed him for a lifetime.”

We actually see the same thing...in software.

In software?

Well, I'm a big proponent of knowing the big picture ("We want to accomplish this goal.") but attacking it through building some really great core technology ("We need to build X and Y, but Z and W we already have.")

In the subprime world, the "technology" that was needed to enable more affordable housing was really great jobs for people who haven't previously had really great jobs. The Bush administration managed to split the rich and poor even further than before--zero wage growth for anybody but the rich in the last 8 years. The investment in the "bottom-up" piece (getting people better jobs) was non-existent, and you could say, regressive.

Even if one was so bald-faced as to say that the end goal was to enrich people, it should have been done by giving them more opportunity, not by giving them more liability.

In software, things that get made this way ("just ship the prototype") tend not to even launch, because they fall apart in years of shipping late, bad performance, and tangled bugs that nobody can figure out.

In effect, projects built on shaky foundations just don't work. Investors in such projects don't get a lot of their money out. Where have I heard that recently?

The dot-com analogy is pretty amazing here, and it's been said that Wall Street had to "invent" something to continue the huge speculative returns they'd been getting in the late 90s. But technology is by its nature somewhat speculative: it works well if you try a bunch of things and only a couple succeed. Mortgages, not so much.

Probably I've taken a weak analogy too far, so I'll leave it at that.