|
Archives
Latest Entries
Links
  |
Atlanta native, academic capitalist, technology professional, spaceflight amateur. Stephen Fleming's blog is at home at http://academicvc.blogspot.com/
Thursday March 26, 2009
Posted by: Stephen Fleming at 10:11PM EST on March 26, 2009
 Jeff Foxworthy said once, "If the directions to your place include the phrase 'Turn off the paved road,' you might be a redneck." I'm not sure if American Maglev would appreciate the comparison, but it's apt. A little ways off Thornton Road in Powder Springs, about fifteen miles west of Atlanta, down an unpaved road, is a glimpse of the future. I got a chance to visit them yesterday. "Magnetic levitation" is one of those delightful polysyllabic phrases. It's routinely invoked in science fiction books and films to connote swoopy futuristic trains hurtling across the landscape at hundreds of miles per hour. But in the pine woods of this Atlanta suburb, Tony Morris (a Georgia Tech graduate) and his team have built a reasonably-swoopy train car and a third of a mile of elevated trackway. A lot of money has been spent here... nearly $50 million over the life of the company. The latest investors are from Dubai. And the darned thing works. Maglevs are cool for lots of reasons. By floating the car over the rail, you eliminate friction. No wheels, no bearings. By inducing eddy currents in the track, the car becomes part of a linear induction motor that's as long as the track. No rotary motors, no gearing, nothing to lubricate. By using electricity as your motive power, you have no local emissions... and if those electrons are coming from nuclear or solar or hydro plants, no emissions at all. And there's not even much noise to complain about. In science fiction, maglevs require superconducting magnets. Superconductors come in two flavors: high-temperature and low-temperature. "High" temperature is a relative term -- these are cooled by liquid nitrogen at minus-320°F, so they're still awfully cold to you and me. But liquid nitrogen is cheap... about the same cost per liter as beer. Unfortunately, high-temp superconductors are pretty lousy from an engineering viewpoint. Hard to work with, brittle, low current densities -- not a good answer. Low-temperature superconductors are cooled by liquid helium. Liquid helium costs more per liter than Glenmorangie. And it boils off and has to be replaced constantly. And keeping anything cooled to minus-452°F means massive engineering challenges. So maglevs based on low-temperature superconductors are horrendously expensive. American Maglev has declared "a plague on both your houses" and isn't using superconductors at all. Just honking big electromagnets (like the ones in your radio speakers, but bigger) with a sophisticated sensor and control system. This system monitors the gap between the magnets and the rail thousands of times per second, adjusting the electromagnet drive current as necessary to maintain a constant 3/8-inch separation. The team's mantra is "smart vehicle, stupid track." This makes a lot of sense. The track is elevated (keeping it out of the path of kids, cows, or cars) and consists of two load-bearing steel rails, two more steel rails to carry the electric current, and a series of flat aluminum plates down the middle. It's supported by pre-cast concrete towers every 88 feet... meaning that building a track means digging sixty deep holes per mile and snapping pieces together, rather than the laborious bulldozer, grading, filling, and paving process needed for traditional light rail (trolleys) or heavy rail (MARTA or Amtrak). Levitating the car only requires 40 kilowatts (50 horsepower) of power; about the same as two dozen hair-dryers. Making it go sucks more current, but you get some of it back when slowing down (by charging on-board batteries). Overall, the company claims to be able to transport passengers and freight for approximately 1/3 the power of steel wheels on steel rails.  Inside, the car feels like a five-fourths-scale model of a MARTA car. It's about two feet wider inside and a bit taller, which makes it feel much more spacious. There's plenty of room for various seating arrangements, or even carving it into compartments if required. The current car would seat 100, or allow 200 to stand at American densities, or allow over 300 to stand at Tokyo density. There's no driver; everything is computer-controlled. For long-haul routes, the company expects there to be an attendant of some sort in each car to make passengers feel more comfortable (and probably to sell snacks and Cokes). For urban routes, we'll probably be on our own -- just like the underground trains at Hartsfield. I've written elsewhere about the advantages of private vehicles over mass transit. But, if we're going to have mass transit, can't it at least be good mass transit? MARTA claims that it can build new heavy rail for $200 million per mile. They're probably wrong; the Los Angeles Red Line cost an eye-popping $700 million per mile in current dollars. (For the math-impaired, that's over $11,000 per inch.) American Maglev claims they can build their system for $20 million per mile. Even if they're optimistic, that's startlingly less than heavy rail. It's even less than light rail... and, because American Maglev's vehicles travel on an elevated track, they don't have to be hardened against collisions with cars and trucks like trolleys... which means the swoopy levitating vehicle will actually be cheaper than old-fashioned electric streetcars. Smart software means that, instead of an eight-car MARTA train controlled by a trained operator, you can have a bunch of single cars running more frequently. I don't know about you, but the statistical indeterminacy of how long I'll have to wait for a MARTA train is one factor in why I don't use transit much. If I knew there'd be another car along every two minutes, I wouldn't care so much. American Maglev's test track in Powder Springs is their third generation of technology, but they intentionally kept a low profile until late last year. Now that they have working hardware, they're hosting frequent delegations of potential customers from around the world. They're bound to win one or more of the projects they have bid on, from Los Angeles to Asia. Personally, I think it'd be great if they convinced the Atlanta Beltline folks to go with maglev! I visited yesterday as part of a delegation from Georgia Tech. I appreciate the time that Tony and his team took to demonstrate the system and answer our questions. We had a great time, and I'm thrilled to see this unexpected new player in Atlanta's high-tech community. Sunday February 15, 2009
Posted by: Stephen Fleming at 2:30PM EST on February 15, 2009
 I've recommended Edward Tufte's books many times... you can see some of my previous comments here and here. He's going to be in Atlanta next month. If your job consists of presenting complex information on paper or on screen, you should go: <http://www.edwardtufte.com/tufte/courses>. People travel from all over the country (world?) to attend these one-day seminars. If you're reading this blog from anywhere in the Southeast, you have no excuse for not taking advantage of the day! Disclaimer: I have no relationship with Tufte other than having shaken his hand after the Atlanta course a few years ago; I don't make a dime if you sign up. If going to the course isn't convenient for you and all you want is the books, buy them from the links below and I'll make a buck per book.
Saturday January 31, 2009
Posted by: Stephen Fleming at 11:55AM EST on January 31, 2009
There has been a lot of noise lately about venture capital funds in trouble because their investors are missing capital calls. Which, at first blush, doesn't make sense because venture funds are supposed to be "decoupled" from public markets. The key bit of knowledge that surprises many people is that venture funds don't have any money. They're not supposed to.
 Venture funds don't have Scrooge McDuck's money bin (although it would be really cool if we did). They'll typically have a business checking account with just enough money to cover a few month's rent and utilities and travel expenses. The big money, the money that is invested in deals, comes from limited partnerships. The venture firm (I'm simplifying here) is the general partner, and large institutional investors or (very) wealthy individuals are the limited partners, or LPs. When "closing" a fund—which, confusingly, is what you call the initiating event; normal humans with shoe stores or law firms would call this the "opening"—the VC firm collects commitments for a certain amount of capital. But that doesn't mean any money changes hands. Instead, the VC firm issues "capital calls" to the LPs when it's ready to make a deal. This will usually happen late in the diligence process, after the target company has accepted a term sheet. Let's make up XYZ Ventures, a hypothetical $100M VC firm (a bit on the smallish size nowadays, but it makes the arithmetic easier). Let's further hypothesize that this firm got 15% of its committed capital from the state of East Virginia's public pension fund... meaning East Virginia pledged $15 million over the "investment period" (usually five years). Now our VCs, through luck and shoe leather, decide they want to invest $5M in SittingDuck.com. They'll send out an email, followed by a phone call, to the East Virginia Public Pension Fund saying something like "We've identified a promising investment opportunity. We're calling 5% of your capital commitment, which for EVPPF is $750,000. Please wire it to our transaction account within fourteen days." When things go smoothly, all the LPs wire their money the same day, and the money can be turned around and wired on to SittingDuck later that day, or the next day. At the worst, the venture fund sits on $5M overnight; the next day, the transaction account is back to zero. Because venture funds are not in the business of holding cash... they're in the business of holding equity investments. VC firms are far too specialized (and their expense ratio is far too costly) to manage cash. Every day that cash sits in their transaction account earning 2% is a day that it is not in a private company earning 25% or more. (Or zero, but that's a different subject.) Limited partners measure VC firms not only on their cash-on-cash multiple ("ten-baggers" and such) but on their rate of return, measured from the day of the capital call to the day of the ultimate distribution (after a sale, merger, or IPO). Sitting on cash earning 2% would inexorably drag down the eventual IRR for a particular deal, and eventually for the entire fund. Which is a longwinded explanation of why VC firms don't have any money, and why they practice "just-in-time" capital calls. Now, firms are getting worried about their institutional investors defaulting on these calls. You can read samples here, here, here, and here. But I haven't seen anyone mention a critical element of this problem, which is asset allocation. Most institutional investors (pension funds, endowments, family offices, etc.) will allocate their investments among a variety of asset classes: stocks, bonds, real estate, hedge funds, venture capital firms, and more. Venture capital (and hedge funds) are categorized as "alternative assets." And most investors are going to allocate a relatively small percentage of their holdings to alternative assets. In many public pension funds, the maximum percentage is written into law. So, to keep the arithmetic easy, let's say East Virginia's public pension fund is a $10 billion fund. (East Virginia is a small state.) And let's say that, since they are more enlightened than Georgia, they are allowed to invest up to 5% of that into alternative assets. They've been doing this for a while, so on August 31, 2008, they happened to have hit exactly 4.50% allocated, or $450 million. That allocation includes their $15 million commitment to our XYZ Ventures. Let's keep making the arithmetic easy: they have 0.5% in cash, and 95% in a mix of stocks and bonds, including professionally-managed funds.
Then the bottom falls out of the market in the fourth quarter of 2008. You've read the papers. The major part of EVPPF's portfolio takes a huge hit. If they're lucky, it's 20%. If they're not so lucky, it's 40%. Let's pick 30%. What has happened to the value of the 5% in alternative assets? Venture capital funds (and hedge funds, but that's beyond the scope of this already over-long post) are illiquid. You can't look up their value on NASDAQ. So the valuations are whatever the fund managers put in their quarterly reports. I'm writing this on the last day of January. Investors are just now getting their 4Q2008 reports. But, honestly, those valuations are going to look a lot like 3Q2008, and 2Q2008. Because VCs tend to adjust valuations of their portfolio companies only when there is some sort of transaction event. Initial valuations are at transaction cost—the price per share that the VC firm paid for them (less a discount for illiquidity). Obviously, a sale or IPO will trigger a re-valuation to the new price per share. Next best, a new venture round at a higher price with new independent investors will usually be sufficient for VCs to mark up their older shares to the new price (less a discount) in their quarterly reports. In the other direction, if the company has raised money at a lower price (a "down round"), the VCs will have to adjust the valuation of their older shares downward... although various forms of anti-dilution can mitigate the impact. But what if the company hasn't raised money? Or if it has raised money, but not with a new independent investor to set the price? Normally, most VC firms will let the existing valuations ride, showing no gain or loss in the quarterly report. Only in the case of a major problem with the deal—the CEO leaves, or a customer cancels a contract, or a product release slips substantially—will the VCs mark down their holdings in the absence of a transaction. And that doesn't happen as quickly as it should sometimes. It's not unheard of for a portfolio company's valuation to be held at cost until it's written down to zero when the company shuts down. So the net result is that the valuations of the illiquid holdings in a private equity firm's portfolio don't fluctuate very rapidly. Whereas the valuations of stocks fluctuate daily. Lately, most of those fluctuations have been down. Back to East Virginia. On August 31, they had $450 million in alternative assets; let's hypothesize that a combination of down rounds, shutdowns, and markdowns have eroded that valuation by 10% on December 31, 2008. But we've already said that the rest of EVPPF's portfolio took a 30% loss over the same period. Do the math:
So what has happened here? EVPPF started off the fourth quarter with an asset allocation of 4.5% in alternative assets, comfortably within their statutory cap of 5%. We're assuming they didn't make any new investments during the quarter, nor did they receive any distributions. But the carnage in the public sector was so great that, without touching their alternative assets, they suddenly account for 5.7% of their shrunken overall portfolio. And, even for money managers with MBAs, 5.7 is bigger than 5. Ouch. Depending on how strict East Virginia's regulations are written, EVPPF may suddenly find itself in the position of having to dump fifty million bucks worth of perfectly sound alternative asset positions onto the secondary market (at a deep discount) just to get below their 5% limit. At the same time, let's say that XYZ Ventures has just found a great new deal and makes a capital call. EVPPF gets the call for another $750,000. EVPPF is still worth over 7 billion dollars, so this call represents 0.01% of their assets under management; peanuts, right? Our hypothetical conservative East Virginia managers are sitting on plenty of cash, so liquidity isn't a problem. But 5.7 is still bigger than 5. It's not that they won't pony up $750K to make their pro-rata portion of the capital call... it's that they can't. Which means that the XYZ Ventures—remember, those guys who try to carry a balance of zero?—don't have East Virginia's money to invest. Surprise! If you're the entrepreneur, you have to hope there are enough investors in XYZ Ventures who don't have strict limits like this to fund an adequate capital call. But, even private pension funds that don't have statutory allocation caps are still going to have internal targets. And 5.7 is still bigger than 5. And that situation is going to continue until (1) the bulk of the portfolio value improves, or (2) until the VCs mark down enough valuations and shut down enough deals to get their proportion of LP assets back in line. The first is completely driven by the public market; have you checked the S&P 500 lately? The second... writing down valuations makes the VC look decisive in the face of crisis, but it's really going to play hell with their chances of raising money for their next fund. (And, just like Congressmen are always thinking about the next election, VCs are always thinking about their next fund. Always.) Shutting down troubled companies makes the VC look equally decisive, but it eliminates the future requirement to keep investing in subsequent rounds of those companies. It's a chance to clear the decks. And, heck, there will always be other companies to invest in, right? Plenty of fish in the sea. To those CEOs of venture-funded companies who aren't yet cash-flow positive... just because you're not paranoid doesn't mean they're not out to get you. Tuesday December 23, 2008
Posted by: Stephen Fleming at 10:54AM EST on December 23, 2008
 It's Christmas Eve Eve, and things are quieting down in the Atlanta technology community. Even the Twitterstream has slowed a bit. Entrepreneurs may be willing to work 24/7, but when all your clients have disappeared for the holidays, even the most dedicated entrepreneur may look up to find some spare time. Which means it's a great time to focus on those event applications that you were always going to get around do someday. Why not today?  Startup RiotStartup Riot 2008 was a blast so we're doing it again! This is a chance for 70 startups (or startup-wannabees) to make a three-minute pitch to 300 people who might be able to help with funding, talent, customer relationships, or other connections. You need to follow the rules that Sanjay Parekh, maestro of the event, has laid out... but this was one of highest signal-to-noise ratio events of 2008. Don't get left out! The 2009 Riot will be February 18th, and Sanjay is already making decisions on applications. Apply to present here, or if you're not currently pitching a startup, apply to attend here. Hurry!  TAG Top 40The Georgia Technology Summit will attract over 700 people to the Cobb Galleria on March 3. Given that both Tom Friedman and Ron Clark will be speaking, this might sell out... sign up now! But, to me, the best part of the Summit is the "Top 40." A selection committee (I'm a member) will choose the 40 most innovative technology companies in Georgia based on specific criteria including degree of innovation, scope and financial impact of innovation, likelihood of success, and promotion of Georgia's innovative efforts nationally and internationally. Eligible companies will be Georgia based, and be viewed as "technology" focused (in other words the companies primary focus would be the development and dissemination of technology). From the Top 40, the selection committee will also choose the Top 10 Most Innovative technology companies, which will make three-minute presentations to the Summit's 700+ attendees at the event. The Top 40 companies will participate in an exhibition at the Summit, expected to draw the state's technology leaders. We'll start reviewing as many as 200 applications right after the first of the year. If you think you have what it takes, apply here before February 1.  Run it By the ProsThe MIT Enterprise Forum of Atlanta has organized their "Run it By the Pros" event for several years now; three companies will get a chance to present their business plan to a panel of experts. I've been on the panel a couple of times, and the companies seem to get a lot out of it. The next one is January 7th, so time is getting tight! Email Virginia@mitforumatlanta.org for more information. So there you have it! Three great events, all open to Georgia entrepreneurs. Put down that premature eggnog, follow the links, and make sure you apply in time! Monday December 1, 2008
Posted by: Stephen Fleming at 5:05PM EST on December 1, 2008
Okay, Mike Schinkel has been assimilated. Good work, all!
I sent him a list of "must-have" iPhone applications, he tweeted about it, and now people are asking what's on my list. I don't claim any special insight or wisdom here, but having spent more than the purchase price of my iPhone in the iTunes Apps Store over the last five months, this is what works for me.  First off, my first Home screen. (Which, in case you didn't know it, is now easily accessible from any of the nine possible Home screens... press the "Home" button once to get to the grid of applications, then press it again to go to the first screen on the grid. So this is where you should put the stuff you use most often.) Apple standardSeven of my "first Home" applications, plus two on the menu bar at the bottom of the screen, are Apple standard: Text, Photos, Maps, Camera, Phone, Clock, Settings, Email, and Safari. Enough said about those. Ten more of the standard apps didn't make the cut, and they're relegated to another screen: iPod, iTunes, App Store, Stocks, Contacts, YouTube, Notes, Weather, Calendar, and Calculator. The first four of those I use only rarely; your mileage may vary. Contacts usually gets launched from the Phone application, and YouTube usually gets launched from within Safari, so they don't get top billing. The last four have been replaced by more capable third-party apps (see below). Free Apps
Commercial Apps
eBook ReadersI have found myself reading a lot of eBooks on my iPhone. I have four readers installed:
All four have advantages and disadvantages. Stanza has the most responsive developer. eReader is the only choice for encrypted books from eReader.com and Fictionwise.com. I could probably survive with just those two. BookZ has some nicer control options, and BookshelfLT integrates best with the Baen online library. Other GoodiesOther apps that are too good to ignore, but who didn't fit into my first Home screen—in alphabetical order:
Most of these are cheap or free. So, that's my list. What's yours? Thursday November 20, 2008
Posted by: Stephen Fleming at 3:57PM EST on November 20, 2008
 We held the third running of the Gauntlet last night. As a heads-up, the December Gauntlet will be a little earlier... 4:30 pm. Here's why. We've held three events so far, and the winners have been: September:Winner: Divvs Runner-up: Simatra October:Winner: 40billion.com Runner-up: Neighborex November:Winner: Jumbis Runner-up: Telanon In December, we'll convene a fourth Gauntlet at 4:30 pm on December 17th, with a hard stop at 6:00 pm. We'll pick another winner and runner-up. Those two CEOs, plus the CEOs of the six companies listed above, will convene at 6:30 for a second event. As usual, they'll each give three-minute pitches (with or without slides). But instead of your sometimes-rude dungeonmasters, this one will be judged by members of the TAG Top 40 committee. The winner will get a slot as one of the Top 40 companies, as well an invitation to audition for one of the Top 10 slots to present at the Georgia Technology Summit next March. And this event is open to the public. Unlike the regular Gauntlets, where you have to be prepared to pitch to get in the door, the 6:30 event will be open to everyone. Come see the eight companies battle it out for the Top 40 slot. Listen to the judges' feedback on the presentations. And get a chance to see some of the hottest new companies in Atlanta! Sign up to present at the Startup Gauntlet web site. If you're not presenting, just show up; no RSVP required. And, as a reminder, winners of the October, November, December, and January Gauntlets get an invitation to present at Startup Riot as well... Friday November 7, 2008
Posted by: Stephen Fleming at 8:24AM EST on November 7, 2008
 Alright, I'm actively courting codgerhood here, but I don't like rap. Or hip-hop. Or whatever else the young whippersnappers call "music." But I just spent ten bucks on a rap CD (well, partially rap), and I love it! If you've been to a Georgia Tech football game this season, you've seen the music video for "A Perfect Option" up on the big screen. That's the second video by a pair of recent GT graduates who call themselves The GTGs (after their campus email addresses); the first was "Take the M-Train." (Which I have seen seriously discussed by Gary Schuster, and which took up a good chunk of the College of Engineering advisory board last spring!) I may not be the biggest fan of their musical style, but the kids are undeniably talented. (Listen to some of the other tracks on "Look to your Left, Look to your Right" and you'll see that they really do display some musical ability!) But the point of this album isn't the music quality, it's the lyrics. Anybody who has ever been a Tech student, or known a Tech student (or married a Tech student!), will understand The Ratio: Sir, if you got a daughter And "Take the M-Train" raises valid points about engineering students switching to management: Why would ya take the harder route when there's an easy way? Now, the GTGs have published a CD, containing these and other songs, two of the videos, and some hysterical phone calls for George P. Burdell. If you were ever a Tech student, you'll understand why I almost drove off the road laughing at Track 4! They position themselves as "Music for the Georgia Tech Community," and that's exactly right. Talent and audacity like this should be encouraged. Go ahead, spend ten bucks on their CD. Buy a second copy at a stocking stuffer for a GT person in your life. You'll be glad you did!
|
