Simple or Complex?

By James Kwak

Ever since the financial crisis, there has been an on-again, off-again debate over the right model for financial regulation. On the one hand are those who favor simpler rules—such as a simple leverage limit based on total unweighted assets—on the grounds that they are easier to monitor and tougher to game. On the other hand are those who favor complex rules—such as the Dodd-Frank Act, which has so far generated over 8,000 pages of rules—on the grounds that the world is complicated so we need complicated rules. For the most part, this has been a shouting match over broad principles.

A friend sent me Andrew Haldane’s paper from Jackson Hole a couple of weeks ago, “The Dog and the Frisbee.” (The title refers to the ability of a dog—or a child—to catch a frisbee by following a single visual heuristic, ignoring factors such as the rotational speed of the frisbee or wind currents.) Now we have evidence.

Financial regulations have become fabulously complex in recent decades. For example, Haldane points out that since Basel II allowed banks to use internal risk management models for calculating their risk-weighted assets and capital, capital regulation is now performed by models that potentially include millions of parameters that must be estimated (p. 9). But these parameters must be estimated using relatively short historical samples—drawn from a historical period that may or may not be representative of the future.

Here we collide with another fact of statistics: when you have a limited amount of sample data, simple models have greater predictive accuracy than complex models. Haldane cites examples from a number of fields. For example, to predict the winner of a tennis or soccer match, you are better off going by the name recognition of the player or team rather than by the quantitative rankings put out by ATP or FIFA (p. 5). And when determining your asset allocation, you are better off dividing your assets evenly among all asset classes than by doing mean-variance optimization using historical data (p. 6).*

So what happens with financial regulation? Haldane analyzes various data on banks from before the financial crisis and on which banks failed (or would have failed absent government intervention) during the crisis. The conclusions? Looking at banks with over $100 billion in assets at the end of 2006, risk-based capital ratios fail to predict which would fail when the crisis hit; simple leverage ratios without weighting assets were statistically significant at the 1% level. (See Tables 3–5 for more specifications.)


The same does not hold for smaller banks, as shown by a sample of FDIC-insured banks. But that’s because of the large sample size. Haldane simulates what would happen if you estimated your models using a partial sample and then tried to predict what would happen to the rest of the sample. In that case, again, simple metrics perform better than complex models.

So what can we conclude from this? One possibility is that banks are just good at gaming the system, and the more complicated the system, the more opportunities for gaming there are. So should we vastly increase the number of regulators to keep pace with the banks and their law firms? That’s unlikely to happen in today’s environment of (self-imposed) budget constraints. More importantly, however, the real problem isn’t that risk management models can be gamed; it’s that they don’t work when the complexity of the model dwarfs the available data from which the model is estimated.

One possibility is increasing supervisory discretion. The risk there is that it would only increase the risk of (and returns to) regulatory capture. At the end, Haldane comes back to the root of the problem. The problem is complexity, and so we need to reduce complexity. One way is to tax it. Another is to insist on structural reforms to complex financial institutions. Until then, all the models in the world will do little more than give regulators the false impression that they are in control.

* Of course, if you’re predicting past performance using past data, there are superior strategies. But that’s not the point of investing, or of financial regulation.

24 responses to “Simple or Complex?

  1. Paging Kerkowski, Mr. Per Kerkowski, white courtesy telephone please: Mr. Kwak got your memo.

  2. Enforce criminal laws before civil regulations, which result in fines and slaps on the wrist. You could indict these banks on money laundering, conspiracy, fraud, theft, collusion. The real teeth is in criminal statutes and penalties, as opposed to some SEC sanction.

  3. @James tells: “A friend sent me Andrew Haldane’s paper from Jackson Hole a couple of weeks ago, “The Dog and the Frisbee.” (The title refers to the ability of a dog­, or a child­, to catch a Frisbee by following a single visual heuristic, ignoring factors such as the rotational speed of the Frisbee or wind currents.) Now we have evidence.
    What the dog is working with, which today’s banking ignores, is the power in life’s, and the computer’s reasoning-ability to record a reusable history. The border collie ignores the Frisbee probable behavior in favor of his own creation of a real-time model of the Frisbee-behavior’s solution. The Frisbee, of course, is taking into account of its own rotation, as well as the wind’s influence in generating Frisbee-behavior.

    Haldane sticks like glue to the probability model, never once mentioning banking’s roots, seven or eight centuries ago, in the double-entry book-of-accounts (See Raymond de Rover’s “The Rise and Decline of the Medici Bank). Double-entry bookkeeping records the bank’s history in a real-time model | control rationale.

    Einstein warned his fellow physicists and mathematicians of the dangers in substituting probability modeling for a real-time statement of facts. As John von Neumann studied neurology in the 1950′s in order to get a better grasp of a computer’s optimum design, he found the existence in reasoning of an analogical number system, versus a digitally coded representation of numbers and words in a binary grammar. Grammar lets the physical model be communicated with great speed and translated back into its analogically modeled physical states.

    A proper double-entry book-of-accounts uses computation’s analogical number system to interpret a real-time — measured — model of current events. Simultaneously — as the ‘double’ in ‘double-entry’ — the book-of-accounts uses the digital grammatical language to compile control-reasoning, as the history of the analogically changing physical model. In the reuse of the digital story-form, the story of behavior is translated back into a real-time analogical model of that behavior.

    The boarder collie is a better Frisbee athlete because he, or she, is not wasting valuable time with probabilities that will complete the solution of the Frisbee-modeling problem after the Frisbee has hit the ground.

    Some interesting persons that, like Andrew Haldane, seldom, if ever, use the word “bookkeeping”: Ben Bernanke, Timothy Geithner, Larry Summers, Alan Greenspan, Simon Johnson, James Kwak. The list goes on and on. If the word appears in a book written by an economist, or a banker, it tends to appear in anecdotal form most often in a reference to bean-counting.

  4. Now Per, don’t tell me you got too much of this:

    Bless your heart, you hell raiser.

  5. Haldane’s paper describes a learning system constantly adapting to reality through accurate observation of reality and application of relatively simple but understandable responses to it, over and over again, Dan. Bookkeeping is a vital part of ensuring an accurate data stream (and as you argue create transparent traceable financial/manufacturable products), but what humans do with that data is a totally different subject. This paper is about managing risk robustly in a regime that is comprehendable by and manageable by humans, and it is at its core arguing for humility; for the restoration of human judgement in regulation but for the design of regulatory systems that focus on robustness to failure rather than the optimization of investment distributions as Basel does now. An all-seeing, all-knowing, all-storing global computer system containing all the market transactions (in a double-entry bookkeeping format, yes) would be just as opaque as the probabilistic internal accounting systems now employed; we would just be giving over our economic sovereignty to another group of high priests. What Haldane is arguing for is a nimble system that – yes, in the process of eliminating questionably founded financial “products”, focuses on the right sort of risk regulation – robustness to failure. This is an excellent paper.
    Einstein DID warn his colleagues against employing probabilistic physics. But he famous admonitions against it have not diminished the power of contemporary physics to provide a working model of the world that allows us to manipulate it effectively. Thermodynamics is probabilistic because it reduces an unimaginably complex particle tracking problem to a series of simple, broadly applicable, robust laws of behavior. It also works because those who employ it understand and work with its limits. So too the theories of quantum physics and their operators.
    The problem with financiers and financial regulators is that they have forgotten the limits of their own models.

  6. @Bond Man, “…Enforce criminal laws before civil regulations, which result in fines and slaps on the wrist. You could indict these banks on money laundering, conspiracy, fraud, theft, collusion. The real teeth is in criminal statutes and penalties, as opposed to some SEC sanction….”

    A world completely disconnected from cause and effect….one long moral and ethical freak show….and I don’t care if I sound like Yoda :-)

    cut and pasted from the article:

    “…Doherty was an “extremely active” member of the advisory board of the Military Religious Freedom Foundation (MRFF), an advocacy group that fights inappropriate religious proselytizing inside the armed forces, said founder Mikey Weinstein, a retired Air Force lawyer.

    “He confirmed for me how deeply entrenched fundamentalist Christianity is in the DoD Spec Ops [Department of Defense Special Operations] world of the SEALs, Green Berets, Delta Force, Army Rangers USAF … and DoD security contractors like the former Blackwater,…”

    Who is stealing the fruits of thousands of years of civilization while a “religious” war is being created by GREED? Christians and Muslims killing each other….?

    Abort the Federal Reserve Board and you abort Armageddon.

  7. @oregano – the only *risk* to reality is an “economy” that is completely disconnected from a sustainable man to land ratio which also does not recognize that there is a LIMIT to profit taking. We are SO there….

    Unless the interest owned by human slave labor to the god-banksters on the 700 trillion in derivatives floating out there in the unimatrix is going to come via the mothership stocked with a new protective atmosphere and ocean filters and a restock of useful extinct species, keeping up the 700 trillion delusion is NOT a “risk”, in your opinion?

  8. @Annie, I never said that it wasn’t…just that the paper is a good argument for humility, simplicity, and transparency. We’ve got a dysfunctional market model, but papers like this one are a hopeful sign that those that control its levers are recognizing it.

  9. The financial sector should not be allowed to engage in the development of instruments and trading platforms that do not improve their efficiency in performing their primary function of allocating resources to overall economic benefit, where they’re doing an absolutely abysmal job, as analysis clearly shows. Instead, they are just enriching themselves. Please see my article guest-authored for OECD at

    Also, high speed computerized trading should simply not be allowed unless the markets are willing to provide the SEC sufficient computer firepower to make sure that the markets are not gamed, or more importantly, not subject to cyber attacks. Otherwise, the potential supposed gains in liquidity just are not worth the huge risks. Just imagine the market impact, especially after the Knight fiasco, if the SEC were to call and tell the Treasury Secretary that “we think that there’s a cyber attack on the market, but we’re not sure.”

    OECD Insights ran a series of articles celebrating the Turing century, to which I contributed, and the one by Dr. David Leinweber of Berkeley National Lab’s Advanced Computational Group addresses the ‘flash crash’ and security issues directly. Please see: and

  10. The desired outcome is reduced financial systemic risk. Simply re-implement basic Glass-Steagall concepts and limit the size of commercial banks to a sensible level. 8 pages of regulation could achieve what 8,000+ pages of Dodd-Frank never will.

  11. None of this matters too much now. As some of us have feared, the banks are now stealing the very land beneath us…think about the implications of this latest edict from the Fed, if you will.

    Fed OKs banks as landlords —
    Lenders get green light on renting out REOs

  12. I think one must first reign in the military, once they get into a disinformation campaign, everything gets lost in the sea of numbers. And the blame for the consequences will lie directly with them.

    Is it beginning to make some sense?

  13. There is a lot to be said for banking as a public utility. Protect them from competition, guarantee banks a reasonable profit, but also force them to adhere to a defined business model. Tell banks what they can do, and that they may only do that, may be the simplest solution of all.

    Prescribe, don’t bother trying to proscribe, you’ll always fail.

  14. @Carla, According to “teacher” Marcus channeling the god-bankster (and one can see why people are asking for a way to measure the impact of a teacher on the students), unemployment won’t get to 7% until 2015 – do people have any idea how much more “land” is going to be stolen by then?

    And the way both commentators sneer at “what people know” about the majik printing press…makes you want to ______ fill in the blank yourself :-))

    “RUTH MARCUS: Well, if it works — and I think that the theory is sound — it will work eventually, but slowly. So I don’t think it will have an effect on the election.

    We won’t probably see the impact of this for six months or so. It reflects two things, I think, from Chairman Bernanke, his lack of confidence — and both of them might be justified.

    One is his lack of confidence in Congress, because he sees Congress not stepping up to the plate to do the fiscal acts that are necessary to get this really lagging economy moving, and a lack of confidence in the economy.

    And so, this was really in some sense his boldest move yet, because he didn’t put a time limit on it. He said, we will be there until the unemployment rate gets to a more appropriate rate. That’s probably — we’re not going see it at 7 percent, according to the Fed’s own views, until some time in the middle of 2015……but the question is, to the extent that the public understands it, I don’t think they are going to be moved by this pretty much one way or the other….”.

    We’re going to have to get Afghani-tough about home and hearth…as Bond Man notes – the rule of law is GONE, completely and absolutely GONE. He wrote, “….You could indict these banks on money laundering, conspiracy, fraud, theft, collusion. The real teeth is in criminal statutes and penalties, as opposed to some SEC sanction…..”. We have stopped prosecuting CRIMINALS – get it?

    @Oregano – only one way out – we issue a currency free of debt to the Federal Reserve. The FRB can wait for the mothership for their supply of “interest” on 700 trillion…

  15. End of the day, the predatorclass gets supericher, the 99% proportionally poorer. The predatorclass is immunized from prosecution for wanton criminal conduct, – the 99% is ruthlessly oppressed, dislodged from their homes and business’, and sentenced to longterm imprisonment for petty crimes.

    Welcome to fascist Amerika, a kleptocracy where the predatorclass, 1% of the population is given extraordinary largess, and advantages from the government – and the 99% is advised to borrow money from, or move back in with their already suffering and overburdened parents.

    Sadly we can all expect QE4, 5, 6, 7, ad infinitum, because the nation and world we inhabit is a fascist kleptocratic state, owned and controlled by the predatorclass.

    Alter and abolish this horrorshow!!! Burn it all down. Reset! The future couldnt be worse for the 99%!

  16. TonyForester writes: “Sadly we can all expect QE4, 5, 6, 7, ad infinitum, because the nation and world we inhabit is a fascist kleptocratic state, owned and controlled by the predatorclass. Alter and abolish this horrorshow!!! Burn it all down. Reset! The future couldn’t be worse for the 99%!”

    Nice post Tony! You comes right to the point! ‘Kleptocratic’ is a great word; I added it to my spell-check dictionary. I have a suggested definition: A kleptocracy believes in the Ponzi Scheme as a fair and equitable way to redistribute wealth.

    I have a question: If our democracy is at odds with itself at the rate of 99%, set equivalent to 1% of its citizenry, why do we the 99% not fix the problem in the ballot box? Could it be that we the 99% are so culturally dumbed down by the present political state of finance that we simply do not have access to an essential intelligence required to run our own democracy? My suggestion is that we look to small ways right in our own neighborhood to fix the dysfunction. We can then use this valuable forum to report on our small successes.

    My own effort is to construct a computer-driven framework of rules that bookkeeper’s followed for seven centuries. I have seen no example of a proper book-of-accounts having every been programmed into bookkeeping software. The double in ‘double-entry’ is designed to expose the Ponzi Scheme on day one of its history. Ponzi, Madoff, now John Dimon at JP Morgan created, and is creating, an irresistible illusion whose success for the 1% must be more fully documented by the 99% who seek a cultural balance.

    I’ve come a long way toward programming a coded example of Ponzi’s Antidote. My greatest obstacle to its success is in convincing my associates that bookkeepers for seven centuries understood accounting solutions that have been lost today. Even at a dysfunction rate of 99% ‘|’ 1% (read that character ‘|’ named ‘pipe’ as ‘set equivalent to’ as ’99% set equivalent to 1%’) we the 1% resist going back into our own history for simple solutions to our inability to act.

    My suggestion: Ask yourself if you believe that there is a solution to the Ponzi Schemes that permeate today’s financial kleptocracy. Then create of list of potential solutions. Report the best three that you would implement. We might then take a vote among the 99% to see which of the sum of best threes wins a majority vote. That list would start us toward a peaceful revolution in today’s finance.

  17. Norm the double entry bookkeepin Carpender

    With finger tools.

  18. Dramatis Personae
    BF (Barney Frank, Chair house banking)
    WSB (Wall Street Bankers)
    Act I: Phone call, 2009
    BF: Guys, I’m telling you, i know you sent me a lot of money, but we have to do something – and, you must admit, you did screw up..
    WSB: no, it is all the fault of to much regulation – we work so hard, and contribute so much to society, and now we don’t even get our bonuses !!
    and, you want to add more regulation, which is the problem in the first place !!
    BF: guys we gotta appear to do something, so this is what we are gonna do: we are gonna write this hideously complex bill, where most of the work gets done by regulators after we write the legislation. That means all the final rules will be done after negotiations and hearings – and who is gonna be at those hearing ? YOU – not to mention, you can dangle jobs in front of the regulators.
    I tell you its genius. All the idot democrats and lamestream media will play along, pretending that we are doing something, and you don’t even have to lean on the morons in the tea party – they will play their part, screaming their heads off about more regulation, not understanding the whole thing is a charade.
    and the best part ? the complexity of the rules will mean you can finally get rid of all those pesky local banks.
    WSB: Barney, thats great – but will people really fall for something this transparent ?
    BF: no one ever lost money underestimating the stupidity of the mainstream media

    Act II, 2011, aboard a Gulfstream II
    WSB: you know barney, it all worked out exactly as you said – every one thought the goal of dodd frank was to restrain us, when it was merely a smokescreen, and just like you predicted, we were the only ones with the money to send lawyers to all the hearings.
    And just to be on the safe side, we hired a couple of the honest regulators at 6figure salarys
    BF: pass the caviar..what is that i see way down below us ?
    WSB: thats nothing barney, just ordinary people who think they have a say in gov’t

  19. A thousand thanks Dan Palanza. It’s Foresta by the way. Like your approach and ideas and we can only hope that our vote and local activism will someday, someway begin the arduous process of righting these horrible wrongs. Sadly, I’m pessimistic on the peaceful solutions. It would be nice if there were peace on earth and goodwill towards men, but alas this is not the world we inhabit. More disturbing is the grim reality that the predatorclass and predatorclass oligarchs will never relinquish their death grip on the government and economy peacefully. I fear more kinetic approaches will be necessary. But, your way is the best way until there is no way forward. Hopefully wise and courageous leaders will emerge to guide us through these turbulent seas. We can only hope and pray and do what we can as individuals to work toward peaceful resolutions to this horrorshow. But time is ticking. The 99% are in dire trouble. The predatorclass are ruthless sociopaths! One spark can turn the entire situation incendiary.

    Desperate people do desperate things! Take away hope, shut off avenues of redress, allow wanton criminality to go unpunished, oppress the people savagely – and sooner than you know it – all bets are off! In a world where there are no laws – there are no laws for anyone predatorclass biiiiiaaatches!!!

  20. @Tony F – “More disturbing is the grim reality that the predatorclass and predatorclass oligarchs will never relinquish their death grip on the government and economy peacefully. I fear more kinetic approaches will be necessary.”

    There you go, no matter how many “red lines” are drawn, they do not operate by the rule of law.

    Great greed always followed by great violence. It’s inevitable.

    The entire planet is sick and tired of psychotic war/drug/slave lords stealing it ALL. And now 99% have NOTHING to lose if they get “kinetic”…

  21. A couple of interesting models of markets favoring simplicity over complexity come in the private capital markets. One is the world of Angel investing, where the favored approach is a very straight forward share and share alike single class of shares equity structure with a hands on board of directors but few formal investor rights deliniated (in stark contrast to the complex deals of next tier up venture capitalists).

    Another is the finanical structure of a typical publicly held company imposed by the the privately owned stock exchanges. These rules historically mandated simple equity structures (although not quite so simple as a single class of stock – until recently common and preferred with the only two kinds of stock allowed), very low levels of secured debt which in turn was always purchase money debt, a single entity structure for most of the value of the business, and debt financing predominantly through a single class of general creditor bonds with considerable equity cushion (although, again, the modern trend inspired by business schools has been towards more complexity with risk differentiated tranches).

    An interesting proposed complexity busting approch to corporate taxation is to impose a property tax on fair market value of an issuer’s publicly held securities, rather than taxing its income. Major attractions are its reliable ability to collect taxes from successful corporations rather than letting some get off tax free, the simplicity of its calculation from easily available data that is hard to game, its reinforcement of the de facto reality that only public held corporations pay meaningful corporate level taxes, its ability to replace existing corporate tax revenues at reasonable rates, its lack of need to make a debt-equity distinction, and its automatic stabliizer effects in the face of business cycles.

  22. Once again, we/re back to Keynes, who, speaking of speculation, said, “The speculation of modern investment markets has sometimes moved me towards the conclusion that to make the purchase of an investment permanent and indissoluble, like marriage, except by reasons of death or other grave causes, might be a useful remedy for our contemporary evils” (page 160, in the chapter on Long-Term expectations from my falling apart paperback edition of General Theory)
    It would certainly solve the evils of flipping corporations a la Bain/private equity, and force them to ride out the entire consequences of their applying burdens of debt and from said debt paying themselves exorbitant fees, then bailing out to avoid the consequences of their actions.

  23. There is an extremely simple solution to the complexity problem. It can be set out in half a page rather than 8,000 pages. It’s as follows.
    Depositors must choose between safe accounts and investment accounts. Money in the former is instant access, it is not invested and it has taxpayer backing (which shouldn’t ever be needed because the money is not put at risk). That money earns no interest, reflecting the fact that the money is not being used.
    Money in investment accounts is invested or loaned on, thus the money is not instant access, plus the depositor carries the risk, but the money does earn interest (reflecting the fact that the money is being used). That way, it’s impossible for a bank to go bust (absent blatant criminality). Thus virtually no taxpayer subsidy for banks is needed.
    BTW, the above equals full reserve banking.
    As for the idea that the above constraint on bank lending would harm economic growth, that is nonsense because government / central bank can easily create and spend extra monetary base into the economy to make up for the latter deflationary effect.