Thursday, May 6, 2010

Stock Market Undoubtedly Broken By Trading Algorithms

In the past year or so major Wall Street players have turned more and more to High Frequency Trading (HFT) algorithms, computer code which automatically buys and sells stocks based on price movements. These algorithms can trade stocks thousands of times a second and are highly profitable because they are designed to rip off all the other market players.

The strategies of the different algorithms make for a very interesting intellectual exercise as they interact in complex and unpredictable ways, forming an electronic ecosystem. Like all ecosystems, there are predators and prey, parasites and symbiotes, evolution and -- as I bet we will find out happened today -- collapse. Systems experts and market insiders have been warning for months that the algorithms take up too much of the market (somewhere around 70% of all volume is now HFT) and that an unexpected major move could easily cause the market to plunge uncontrollably. The solution is to ban them by making execution time as "long" as a second or two.

Today the market fell around 5% in less than five minutes, most likely artificially caused by HFT in response to the very real stress in the global debt and currency markets. [The Street has identified a few of the individual stocks that broke down.] I happened to be looking at some real time feeds and they were shut down, in fact almost all the major financial sites and blogs were sporadically offline for about 30 minutes. In that time the market rebounded: my guess is that they shut down the trading computers and bid the market back up through floor traders.

The debt problems will last for years and are very complex to address, but the HFT meltdown is simple to fix. Until it is we will experience more occurrences like today and increase the risk of a prolonged meltdown.

Friday, April 16, 2010

SEC Files Civil Fraud Charges Against Goldman Sachs

Wow, I never thought this day would come. It has long been an open secret that Goldman Sachs had a role in defrauding subprime mortgage investors, but due to their specific government connections and the general state of the financial sector it was assumed nothing would come of it. However the SEC has filed civil (for now?) charges against Goldman to recoup their ill-begotten gains.

This could have far wider implications: for instance more details are emerging about a hedge fund named Magnetar that helped fuel massive demand for subprime mortgages in order to short them. Yves Smith claims that Magnetar alone accounted for 35-60% of subprime mortgage demand in 2006. She also details their financial contributions to Rahm Emanuel.

Bill Black, the lead regulator in cleaning up the Savings and Loan crisis, has long stated that fraud is endemic in our financial system, particularly in the mortgage arena. In this five part interview he explains how he believes hundreds of billions of dollars of fraud occurs and charges the megabanks with facilitating this at the highest levels. Congressional hearings on Washington Mutual support this view. In his expert opinion, the financial crisis is not as much about misguidance and ignorance as it is about willful criminality.

It is obviously too early to tell where this case will go, let alone whether it is a precursor to more suits, but hopefully it is the beginning of accountability and a necessary step for true economic recovery.

Wednesday, April 14, 2010

Does Multi-National Corporate Success Mean Anything To US?

I don't want to harp on Dorian's post of optimism (OK I do slightly) but I do find it telling that Daniel Gross' breezy article argued for US economic success made up nearly entirely of anecdotes about major multi-national corporations. It is becoming less  clear whether success of these corporations really helps the USA in many ways other than the stock market.

Mish has a great post on this (even if you don't agree with his political slant). Much has been made of labor being moved to overseas for wage reasons, but taxes are very important too:


In 2004, U.S.-based multinational corporations paid about $16 billion in U.S. taxes while earning about $700 billion offshore, an effective tax rate of about 2.3 percent, according to the administration statement...

Clinton administration officials realized they also had made it easy for multinationals to create entities whose only purpose was to shift profits into low-tax countries and out of reach of the tax authorities, according to a January Government Accountability Office report that found 83 of the 100 biggest companies had subsidiaries in tax havens.

Once the assets were in the haven, the U.S. parent company borrowed from the subsidiary. The interest payments were deductible in the U.S. and tax-free in the haven, the GAO said. The nonpartisan congressional Joint Committee on Taxation recommended in 2005 that the rules be repealed.


And:


As you work on your taxes this month, here's something to raise your hackles: Some of the world's biggest, most profitable corporations enjoy a far lower tax rate than you do--that is, if they pay taxes at all.

The most egregious example is General Electric (GE). Last year the conglomerate generated $10.3 billion in pretax income, but ended up owing nothing to Uncle Sam. In fact, it recorded a tax benefit of $1.1 billion.

How did this happen? It's complicated. GE's tax return is the largest the IRS deals with each year--some 24,000 pages if printed out. Inside you'll find that GE in effect consists of two divisions: General Electric Capital and everything else. The everything else--maker of engines, power plants, TV shows and the like--would have paid a 22% tax rate if it was a standalone company.

It's GE Capital that keeps the overall tax bill so low. Over the last two years, GE Capital has displayed an uncanny ability to lose lots of money in the U.S. ...

It only makes sense that multinationals "put costs in high-tax countries and profits in low-tax countries," says Scott Hodge, president of the Tax Foundation. Those low-tax countries are almost anywhere but the U.S. "When you add in state taxes, the U.S. has the highest tax burden among industrialized countries," says Hodge. In contrast, China's rate is just 25%; Ireland's is 12.5%.

Corporations are getting smarter, not just about doing more business in low-tax countries, but in moving their more valuable assets there as well. That means setting up overseas subsidiaries, then transferring to them ownership of long-lived, often intangible but highly profitable assets, like patents and software.


Large corporations never did have much hiring during the last expansion, as they focused primarily on international growth. The majority of jobs and nearly all of the job creation is now at small businesses, which by all accounts are doing terribly. Multi-Nationals may indeed produce new innovation and products as Daniel Gross suggests, but I think he overlooks the reality of the situation: nearly all the materials, manufacturing and increasingly the design will be done overseas. In fact a lot of large companies are "American" as much for political protection and benefits as much as anything else. And for that, we see a pittance in tax collection.

The companies where this is not an accurate characterization (e.g. Google and the like) not only have few employees in the scheme of things, but will increasingly move overseas. As Mish's reader pointed out, good old IBM has reduced American worker headcount by 30% in the last five years and now has 70% of its workers in foreign locales. Low economic growth in the US and increased taxes will exacerbate this trend unless we change other policies.

Is Our Economy Back? The Record Downturn Makes It Hard To Know

Dorian, the site's (increasingly less lonely) optimist wrote a post earlier linking to Daniel Gross' article in Newsweek.

I have long been skeptical of "green shoots" talk as being bad statistics; for instance the economy grew 5.9% in the fourth quarter, but after inventory adjustment is stripped out the increase was only 2%. There are many measures showing that inventory adjustment is nearly complete and we will lose that contribution starting in Q2 or Q3 at the latest. I argued that GDP is constructed in such a poor way that it is nearly a useless statistic at this point in time, and that instead it was far better to look at the individual components. I still stand by that assessment, but the water is getting very murky on those points too.

Consumer spending increases are a must going forward and there was a great report released for March. In fact overall spending has been very strong in face of all the adversity we are seeing, something that has surprised me.

However, against that backdrop other things don't make sense at all. For instance small businesses are really hurting a ton and are near record lows in their optimism, even tough traditionally they have been the first to benefit from an expansion. Meanwhile, Gallup's self reported discretionary spending poll has been flat since Feb 09, seasonal events excepted. And lastly, state sales tax collections are abysmal, often around the same levels as last year.

How can we reconcile the topline reported increase in sales with all the other data? Well if you look at the methodology of the sales report, you will find that it estimates sales based on same store sales excluding price changes. As Mish points out in the prior link, many chains have gone bankrupt and many existing chains have severely cut back on the number of stores. Thus, if people are spending a similar amount of total money as last year then same store sales will increase and the consumer spending release will rise. Moreover, the small business survey reports very poor pricing power and a lot of price decreases. Again, with minimal total inflation and an increase in buying items on sale or with coupons, the total can not change very much but the official report will show increases.

The board responsible for dating recessions recently said it was too early to call an end to the recession. That doesn't mean much, they often wait for a year and a half after a recession is over to date it. However, it did prompt Calculated Risk to have a great post showing the measures they use. I STRONGLY encourage everyone to check it out.

Not only does the depth of the downturn really shine through, but the dynamics of the recovery do as well. GDP fell a lot, but is now increasing at a historical rate, however income is lagging strongly. Industrial production fell a ton and is making a solid but under performing recovery. Employment is of course a mess, and income less transfer payments has made no recovery at all. It is very clear from the last two metrics that the government is the only thing holding together the economy at present.

I have a one word take away message: confusion. We are in such an ahistorical period that the entire way of looking at the economic world has changed and any argument can be made going forward. This is why banks can make record profits and yet still have trillions of dollars of losses; GDP can increase 6% but see employment losses; income can fall 7% sans transfer payments, but retail sales have been (or maybe haven't been) strong. People can make nearly whatever argument they want as a strong possibility going forward, from complete collapse of the world economy to the strongest recovery in history, and the data will support it.  It is no longer just about dollars and cents, but about mindset.

I personally feel that the risks are still very much strongly on the downside and that we will suffer from a surfeit of optimism instead of a deficit. I think that the statistics that show a strong recovery are based on flawed assumptions, and a sense of complacency will harm us going forward. The government is continuing to prop up the economy but it seems like all sense of urgency about making structural reforms as been lost, and that companies are more obsessed than ever at short term gain over long term wisdom. But that's just a feeling, hopefully I'm wrong...But that's just a feeling, hopefully I'm wrong...I've felt this way for quite a while now and it hasn't occurred, so maybe one day I will wake up and decide I was completely led astray.

Tuesday, March 30, 2010

One Last Thing

I'm not good at goodbyes (I literally have to make a conscious effort to say goodbye to people in order to not offend them) so in case comments and/or commenters don't come back: uh goodbye.

OK now with that out of the way I want to take advantage of the temporarily touchy feely environment for my own nefarious purposes. If you don't mind could you leave a comment stating your profession and/or skillset/interests either here or on my cross linked blog. I say this because in reality the intent of my entire blogging career was a big long setup to make connections that I could then utilize for real world projects that I am kick starting shortly and am seeking interest. Thanks.

Monday, March 29, 2010

Internet Comments As Tragedy Of the Commons, Franklin Paradox, Society Writ Large

I am very happy to blog for TMV because I have cultivated a few online acquaintances and more importantly it makes me feel important to imagine thousands of people hanging on my every word. But yeah, pretty much the only reason I write posts is for comments, so I'm sad to see them go and which is why I am going to cross post everything at a blog with the sole purpose of having comment threads (http://mikkelattmv.blogspot.com/). A link to the post will be at the end of every post at TMV and I hope that at least some of you follow it to comment.

That said I don't begrudge the decision and have a few things to throw out there. First of all, I think this is a classic example of tragedy of the commons. We all (including me) have benefited from the work of a very few and the disconnect between the effort and reward leads to changed behavior. I also think that the reaction on the post announcing the change shows the Franklin paradox. "The Ben Franklin effect is a psychological finding: A person who has done someone a favor is more likely to do that person another favor than they would be if they had received a favor from that person." Instead of relationships being close to zero sum quid pro quo, in reality they tend to default into favor-receiver relationships where the favor side feels obligated to continue providing and if at some point they decide not to then the receiver side will be upset that they lost something they expected. In all the comments on the prior thread, only a few every had any shades of thanking the moderators that have spent countless hours to keep the site civil thus far. This isn't an accusation, just an observation. Even though I am saddened by the decision my primary impulse is to give a big Thanks to the site moderators, especially Dr. E, because ever since I found out how much work she puts into the site I knew that it was allowing me to get something out of it I wouldn't otherwise.

To me these two templates really go a long way in describing the social framework of what has happened and I (perhaps naively) look for them as a source in which to look for future correction. For instance work done on the tragedy of the commons has shown that the outcome is very dependent on the source of governing authority. When authority is centralized into a hidden source that passes down the rules then it tends to lead to negative outcomes, while if authority is localized and created amongst the people that use the shared resources then it tends to lead to positive outcomes. A recent Nobel winner used historical examples to build models explaining how decentralized authority with shared broad intent will lead to the best outcomes. When it comes to comments, this suggests that instead of having the burden be on a single individual or small group, that many authors and select long term commenters should share the burden of regulating comments by agreeing to general broad rules of intent and then sharing the task of actual policing.

When it comes to the Franklin effect, the key is to realize that both sides come into play: yes if you are a receiver in a relationship then it is wise and just to extend help when the giver expresses distress, but it's also the giver's responsibility to communicate their viewpoint and ask for help. Some relationships will break down because the receiver just wants to be a leech and that's OK, in some relationships the receiver actually feels somewhat guilty about their role but doesn't know what to do (or believes that the giver likes the job and would be upset if they couldn't do it) and in some relationships roles will switch. It's hard to know a priori, the key is to have open communication and see if resolution is possible. Only a few comments even hinted at valuing the comments enough to put any work into it (and in general I try to offer contribution if I'm the receiver and something has changed) but I also feel like TMV editors should have reached out sooner.

I do think that internet communities are good analogs of "real life" communities in many ways, and have noticed that comments are a particular sticking point that really mirror it. In general sites that are more general audience have a much harder time than sites that have arcane interests -- there aren't too many trolls that will visit 19th century basket weaving blogs. Sites that have explicit institutional order with different commenter roles and advancement up the ranks to moderator tend to form larger and longer lasting communities than those that don't have this. Sites that have high purity tend to get more participants but at the tradeoff of losing interesting ideas and real engagement.

Anyway I want this post to sound pretty neutral and more observational and will leave it to you guys to say what you will. I look forward to having my delusions of elegant explanation torn to shreds once I'm the one that has to deal with the real life ugliness of moderation...and really appreciate the time that everyone at TMV has spent to shield me from that (and web design, hosting, so on and so forth) thus far and any work in the future.

Two Amazing Graphs About Consumer Spending

Calculated Risk has a post that must be seen to be believed. First off, look at personal consumption in February, which is back at pre-recession highs. This is quite amazing, and I admit, I am completely flabbergasted. No way did I think that consumer spending would rebound and that is a big assumption in my belief that we are going to have a double dip recession (well more to the point it still hasn't ended) in the near term. This may look positive, but then look at the second graph: personal income minus transfer payments, which are direct payments by the government to individuals. That measure never rebounded and is still near the cycle low, even decreasing in February. This measure really captures the importance of government payments in trying to plug the consumer spending graph; it is about $800 billion (~8%) below the pre-recession high.

So what does this mean? Looking over time it is another measure clearly showing that the nature of recessions has changed in the past few decades. Every recession before the 2001 recession had the personal income minus transfer payment measure trough be coincident with the end of the recession and equaled or exceeded post recession highs within one year. The 2001 recession did not start increasing for nearly 16 months after the official end. Of course in that recession the measure was flat instead of dropping precipitously like we have seen. These measurements are demonstrative of the structural imbalances in a labor market that is growing increasingly unable to create jobs; indeed the post 2001 recession saw job creation begin in the middle of 2003, due to being in the middle of the biggest housing bubble in history.

It also means that government outlays are responsible for holding our economy in its current stagnant state, a situation that led to a combined deficit of nearly $3 trillion in the 2009 and 2010 fiscal years. I view it as a high stakes game of chicken with a catch 22 built in. The government cannot keep running such high fiscal deficits for very much longer, and as long as sustainable consumer demand is depressed then ending the high deficits will lead to another major recession; however if the government spends too much time trying to keep the system propped up then interest rates and debt loads will be so high that it will severely hurt us in the coming decades. On the other hand sustainable consumer demand can only grow based on secure job growth but companies won't hire until they see it happening already. In the past increased short term government spending was enough to jumpstart the recovery, an outcome that seems far from certain now.

It reminds me of the most misused medical device on TV: the defibrillator. On the shows it is a magical device that is readily applied to everyone that has no pulse or looks like they are dead, miraculously saving or tragically failing based on the needs of the plot. In reality defibrillation only works when the heart is still beating, just not paying attention to the pacemaker or going through the cycles needed to pump blood. With the same symptom (i.e. no pulse) the treatment will be successful or not depending on state of the heart, if there is no rhythm (asystole) then defibrillation doesn't work. My concern is that by ignoring the underlying state of the economy we are not giving the proper treatment and our current fibrillation will turn into asystole.

As wikipedia states:


Ventricular fibrillation is a medical emergency that requires prompt BLS/ACLS interventions because should the arrhythmia continue for more than a few seconds, it will likely degenerate further into asystole (a flat ECG with no rhythm- which is usually not responsive to therapy unless there is still some residual fine VF rhythm left or the patient is otherwise lucky AND is treated very quickly); after this, within minutes blood circulation will cease, and sudden cardiac death (SCD) may occur in a matter of minutes and/or the patient could sustain irreversible brain damage and possibly be left brain dead (death often occurs if normal sinus rhythm is not restored within 90 seconds of the onset of VF, especially if it has degenerated further into asystole).


To really stretch the analogy we can say that the response so far as been to give CPR by pumping the system with money, and I just hope that the focus turns to getting the rhythm back instead of declaring victory and letting asystole form.