Nicholas Nassim Taleb and Mark Spitznagel have a provocative comment up at the Financial Time today, In some ways, it is isn’t surprising for those familiar with his work on risk and uncertainty. On the other hand, it is an eye opener to see what an internally consistent, reasonably comprehensive solution to our mess looks like,
Taleb and Spitznagel, unlike most others, include real economy fragility in their calculus. The dependence on sophisticated computers networks and advanced communications creates more points of failure. Integrates supply chains and interdependence due to trade also creates more complexity. For the life of me, I never understood the vogue for outsourcing and offshoring, Every study ever done says the majority of companies are disappointed with the results. It seems to represent hope over experience. And from what little I have seen in the way of hard numbers says it does not yield impressive results. One IBM project, to send some work to China, showed that the labor cost savings were substantial, something like a 75% to 80% reducution. Yet the all-in cost savings projected were a mere 15% to 20%. And most projects do not live up to expectations. The gap between the two figures says the additonal coordination costs and delays were considerable.
That is a tremendous amount of rigidity and risk to introduce into one’s operations for not much gain. Yet everyone went merrily down that path because it was what all right-minded modern companies were supposed to do.
The other big message of this piece is trying to prop up asset prices via more debt is a bad idea. He recommends restructuring via large-scale debt to equity conversions, plus arrangements that allow for partial equity conversions if borrowers go into arrears. A clever concept, but too hard to administer.
As much as Taleb and Spitznagel’s message about the dangers of debt, and the need for a surgical remedy is clear, the odds of the right sort of action are zero unless things get much worse in short order.
From the Financial Times:
The core of the problem, the unavoidable truth, is that our economic system is laden with debt, about triple the amount relative to gross domestic product that we had in the 1980s….government policies worldwide are causing more instability rather than curing the trouble in the system. The only solution is the immediate, forcible and systematic conversion of debt to equity….
First, debt and leverage cause fragility; they leave less room for errors as the economic system loses its ability to withstand extreme variations in the prices of securities and goods. Equity, by contrast, is robust: the collapse of the technology bubble in 2000 did not have significant consequences because internet companies, while able to raise large amounts of equity, had no access to credit markets.
Second, the complexity created by globalisation and the internet causes economic and business values (such as company revenues, commodity prices or unemployment) to experience more extreme variations than ever before. Add to that the proliferation of systems that run more smoothly than before, but experience rare, but violent blow-ups.
Our ability to forecast suffers due to this complexity and the occurrence of the occasional extreme event, or “black swan”. Such degradation in predictability should have made companies more conservative in their capital structure, not more aggressive – yet private equity, homeowners and others have been recklessly amassing debt. Such non-linearity makes the mathematics used by economists rather useless. Our research shows that economic papers that rely on mathematics are not scientifically valid. Not only do they underestimate the possibility of “black swans” but they are unaware that we do not have any ability to deal with the mathematics of extreme events. The same flaw found in risk models that helped cause the financial meltdown is present in economic models invoked by “experts”. Anyone relying on these models for conclusions is deluded.
Third, debt has a nasty property: it is highly treacherous. A loan hides volatility as it does not vary outside of default, while an equity investment has volatility but its risks are visible. Yet both have similar risks. Thus debt is the province of both the overconfident borrower who underestimates large deviations, and of the investor who wants to be deluded by hiding risks. Then there are products such as complex derivatives, which in the name of “modern finance” make the system even more fragile.
Against this background, we have two options. The first is to deflate debt, the other is to inflate assets (or counter their deflation with a collection of stimulus packages.)
We believe that stimulus packages, in all their forms, make the same mistakes that got us here. They will lead to extreme overshooting or extreme undershooting. They lead to more borrowing, by socialising private debt. But running a government deficit is dangerous, as it is vulnerable to errors in projections of economic growth. These errors will be larger in the future, so central bank money creation will lead not to inflation but to hyper-inflation, as the system is set for bigger deviations than ever before.
Relying on standard models to build policies makes us all fragile and overconfident. Asking the economics establishment for guidance (particularly after its failure to see the risk in the economy) is akin to asking to be led by the blind – instead we need to rebuild the world to make it resistant to the economist’s mystifications.
Invoking the pre-internet Great Depression as guidance for current events is irresponsible: errors in fiscal policy will be magnified by this kind of thinking. Monetary policy has always been dangerous…. Bubbles and fads are part of cultural life. We need to do the opposite to what Mr Greenspan did: make the economy’s structure more robust to bubbles.
The only solution is to transform debt into equity across all sectors, in an organised and systematic way. Instead of sending hate mail to near-insolvent homeowners, banks should reach out to borrowers and offer lower interest payments in exchange for equity. Instead of debt becoming “binary” – in default or not – it could take smoothly-varying prices and banks would not need to wait for foreclosures to take action. Banks would turn from “hopers”, hiding risks from themselves, into agents more engaged in economic activity. Hidden risks become visible; hopers become doers.
It is sad to see that those who failed to spot the problem (or helped to cause it) are now in charge of the remedy. Just as the impending crisis was obvious to those of us who specialise in complexity and extreme deviations, the solution is plain to see. We need an aggressive, systematic debt-for-equity conversion. We cannot afford to wait a day.
42 comments
Comments are closed.
Taleb and Spitznagel ignore important institutional realities. The key question is who profits from all this debt and leverage? The top investment banks, hedge funds and private equity funds. There are powerful interests fueling the expansion of debt and leverage. The problem is that this course has its limits, especially when debt deflation hits the economic system. Once this happens, you hit a brick wall and it becomes a lot tougher to get out of the rut. This is why central banks around the world are desperately trying to create inflation. All they will do is create another asset bubble, not real economic inflation.
cheers,
Leo
Regarding the meager benefits of outsourcing, I have seen many a CEO balk at giving a raise to a worthy employee barely making minimum wage (whose departure would cripple a portion of the business) and not flinch at a $25,000 couch for a third floor lobby he never even sees and literally has no value for the business at all. There is something about directly paying other humans that upper management finds particularly offensive (other than themselves of course). Outsourcing solves this problem by pushing the money from people to things, even in small ways and even if they are paying unseen people who are making those things (yes, this is irrational; yes, this is my point). This is perhaps not so surprising given that relative consumption has generally been shown to be much more highly motivating than absolute consumption. I don't think there is a grand conspiracy for class warfare, but there are surely small yet ubiquitous incentives to engage in it on an individual basis even if we don't realize we are doing it.
25 percent reduction is the best case what I have heard about outsourcing cases in IT industry. That means already years of experience in Asian countries before starting the outsourcing.
Add to that cultural differences leading to long delays, lost customers because of lower quality etc etc and outsourcing really is not worth the risk in manufacturing sectors requiring engineering skills.
Outsourcing does not have to make sense to the company as a whole. For example, employees generally will not offer bribes to be hired for a position. Contracting firms however will offer bribes to obtain contracts. It can be quite profitable for upper level management to outsource. (Surely a squeaky clean firm like Satayam would never offer bribes, right?)
Amen. I was suspicious of just in time inventory the first time I was exposed to it. We have just in time finance today. Banks which need to buy funds everyday. Thanks reckless. But the banks have computer models which make this practice look safe. So?
"It's the debt, stupid," begin the authors, and I concur completely. Everyone in authority around the world believes that more debt issuance is the answer to any problem—except the Germans, sort of, who are for that reason reviled as chuckleheads by all subgeniuses of the 'more debt' persuasion. At some point, one either converts the debt to equity in some orderly way, or at a timedepth not long after the debt is defaulted on in a disorderly way. All other resolutions have probabilities vanishingly small.
We as a society need a) to raise revenue, which means either b) to raise productivity, c) to redistribute existing revenue, or both b) and c). Redistribution is anathema to our present Powers That Be, so they are all for d) none of the above but more of the same, i.e. more debt issuance. There are arguments about who should issue the debt, and more arguments regarding what the funds raised from debt issuance should be spent on or whom given to, but contemporary economics is near unanimous that 'more debt' has the best outcome. And these subgeniuses are 'the experts' so they know what they're talking about, right?
Debt issuance is like a license to pull nodules from ones arse, paint them gold, and trade them for things of inherently greater value. So of course eveyone likes this proposition since everyone is equipped to play. Paying back that debt is rather less fun, but the way up is so joyous that only a trained priesthood sworn to poverty could restrain the rest. The Fed! Er, well they lost their faith, but economists as a group have yet too since they are still paid to mouth the old shibboleths. Debt is such a gas of a drug that only the reality police keep it under control on historical timescales. Those flashing lights are approaching head on, boys and girls. Prick up yer ears 'n' yew can hear the alarum.
Regarding outsourcing, this was always done as a big Fuck You to workers by management. Seriously, its drivers are socio-political and only pretended to be economic. Once labor got the right to organize legally, managment has always been looking for a way to get out from under the need to 'talk to their workers like people.' It's so, so de classe, what? Management got in the habit of seeing workers as costing them money rather than making them money. Well in outsourcing, they don't have to talk to their workers because their workers aren't even _their_ workers. Of course every firm going to outsourcing overseas has bet their companies existence on things like uninterrupted shipping routes, stable commodity prices, stable currency regimes, and all the rest. Many more layers or risk, and losss of product control. Then they needed to hedge those risks in the financial markets, adding counterparty risk and market manipulation risk (regarding which no one talks enough about 'because markets are perfect and self-correcting' [sic] but there it is). And often, the end product is worse, but if everyone's doing it there's no loss of competitive advantage.
So the net result of outsourcing has generally been to make slightly less good products with considerably more risk and systemic awkwardness, but without any need to treat those workers as real people. And since there were fewer workers but just as much revenue flow that makes bigger slices for management, natch.
Outsourcing had nothing to do with improving productivity or making a better product. It is inherently a an ongoing offensive in a class war regardless of whether it is so conceptuallized by many or any of those who practice it. . . . But that's how we do things here until we don't, so there it is. And the public doesn't want to change that. That vote for Obama was a vote for the good old days of That 90s Show, not a vote for 'change,' and specifically not a vote for a change in the sociopolitics of labor in the US. The chumps still think that management will give them an even break, and if they mean in their femur they might actually be right . . . Instead of just voting center-right.
anon @ 4:19 says:
Outsourcing does not have to make sense to the company as a whole. For example, employees generally will not offer bribes to be hired for a position. Contracting firms however will offer bribes to obtain contracts. It can be quite profitable for upper level management to outsource. (Surely a squeaky clean firm like Satayam would never offer bribes, right?).
That's exactly right, and it's true not only of big corporations or whole industries but of all large structures, including abstractions like "the economy" or a "country".
There's no such thing as one monolithic interest, and no such thing as a policy that benefits everyone. You always have to ask, cui bono?
More often than not, where money, power, or ideology are at stake, the decision will be to the benefit of the immediate decision-maker, even at the direct expense of the so-called whole.
Since there's no way to solve this within the framework of large structures, the only solution is to do away with large structures.
And since we're on the subjects of debt and out-sourcing, it's a good moment to revist the notion of fiscal stimulus. The fiscal stimulus as presently undertaken in the US is, in and of itself, affordable relative to the scale of our economy. It is likely to be even less effective than those who advocate it bill it to be. First, there's the issue of a stimulus domestically being, in part, a 'stimulus for Chinese exporters,' as has been put, accurately, by sum. Significantly more than in the early 20th century spending for consumer consumption leads to export of money rather than the circulation of money domestically. Yes, yes, this is very much all the blathering at the G-20 was about. The powers that be and many economists are aware of this forseeable yet despite that unbuffered systemic problem from the globalization of production. The fact remains that to the extent to which US consumers and smaller businesses actuallly get their hands on any of that stimulus money a substantial though minority slice will sluice offshore, adding little to the velocity of money in the US or to further demand or to available credit.
But beyond that, "It's the debt, stupids." The consumer in the US is greatly endebted. Even many who are employed and in principle financially sound are so exactly because they were sufficiently affluent to have a modest asset base. A base which has been severely damaged by asset declines in equities and real estate at least to the point where the future spending plans of those, too, are duff down the shredder. We aren't faced with 'the paradox of thrift' here, because to the extent to which either set of folks actually gets their checkbooks on any of that stimulus they aren't 'saving' it. They're trying to pay down debt, or unload underwater assets at minimal losses. And paying off debt or moving those assets sends the money into the top of the financial system, which is likely in present circumstances to do nothing to add to employment, stimulate demand, nor support the velocity of money. Sending money back into the financial system in debt repayment repairs, very modestly, their eviserated capital bases; to the extent to which it succeeds in that, it will faciliate speculation. What we have, my friends, is tha paradox of usury, where all money entering the economy gets sucked into the capital-destroying mechanism called the money center banking system.
In and of itself, a stimulus isn't a bad idea. The theory is not unsound, and in the right context it might help stirr up enough money in motion to inch us away from debt deflation. This is not that context. Those pushing stimulus have tried to 'factor up' for the capital and velocity lost offshore. But reading them (here's to you, Paul), I do not get the sense that they fully grasp the debt position of the US consumer, or it's implications, or the implications of pushing money into that problem set. We may get some nudges on the _employment_ front, the one front that really matters, from stimulus, but all but surely much less than projected. And such oomph as is generated will be ghouled up by the money center banks sucking away at the riven body of the real economy.
Which means that after this year's $1T deficit, and the one planned in kind next year we will have just about locked our tiller hand on into a debt/currency crisis. Which is why the advice of the two authors in Yves post is really more relevant than any 'stimulus.' We haven't repaired the banking system, or done jack to fix the severe imbalances in it, nor in any significant way as yet reduced our overall debt position. Instead, the idea is that Uncle Feddy and Uncle Sam will borrow the money to make us good. There are only two sources for that: greater fools ovreseas or our children. The invoice is in the mail to both: let's see who honors the bill and who sends it back faster than it went out. Got an equation for that one, economists? I'm sure that debate will be . . . stimulating.
I think the real evil is not too much debt, but too much work which has to be created to fill a too long official work week.
The entire edifice of economic policy is devoted to creating the necessity for work where none exists – and this has been true since the Great Depression. Debt is merely the way of facilitating this.
"In and of itself, a stimulus isn't a bad idea. The theory is not unsound, and in the right context it might help stirr up enough money in motion to inch us away from debt deflation."
I am afraid – no I am not – this time stimulative spending by government will not work. It is broken, as monetary policy is broken.
I have a theory:
The reason the stimulus plan was targeted to begin in earnest after unemployment peaked (it begins in 2010, but the administration projected unemployment peaking in 2009 Q4) is that they were hoping employment would bottom first, because the damage done to the dollar by this package will be deadly.
They were wrong, and now a tsunami of paper will hit the market and the dollar will plummet.
So…if outsourcing is inefficient today, it will also be just as inefficient tomorrow?
C'mon.
In all of these studies you are reading on the issue, are they also stating "empirically" that not only is outsourcing inefficient, but "we have found no progress over the past decade at reducing these inefficiencies".
It's understandable to take issue with outsourcing for a number of reasons, but the IBM analysis has a huge gap of its own in refuting the outsourcing argument.
I think it was ford who found when they outsourced all their IT that the savings came at the cost of quality and time. The end result was that they brought it all back in again. There is intangible value in having your workers being on the same social wavelength, speaking the same language, and working in the same time zone which always gets missed out of the calculations.
As for banks or more normally social housing associations take a part stake in your house this is already big business in many countries. I don't expect it to really get off the ground in the US because it limits how and when you can sell and usually affects inheritance rights and you still need to pay rent on the part you don't own. Typically this is something that is entered into in attempt to reduce maintenance costs as the home owner can usually get the minor party to part with money to protect their investment.
" For the life of me, I never understood the vogue for outsourcing and offshoring," – umm – unions? can't unionize a job in a foreign country and it's much harder to unionize a us job when the board keeps threatening labor with Asia or Mexico. Keeping labor down means keeping wages down which means requiring those with repressed wages to turn all their equity into debt.
It took a generation to do it but outsourcing has played a not trivial road in knocking the American worker and therefore consumer flat on her/his back. Which would be fine if we didn't have a consumer driven economy.
“…..a tremendous amount of rigidity and risk to introduce into one's operations for not much gain.” – Yves
I love it here. It gives me the audacity to hope that maybe somebody alive can see the real effects. Moving from vertically integrated companies (example GM of the 70’s) to horizontally integrated companies (example Dell of the 90’s) may indeed provide for quicker implementation of improvements, better incentives to make those improvements, and diversify some risks away from single captive business units, but it comes at the cost of an extensive managerial infrastructure and makes production a commodity with margins so narrow that it can carry only the cost of production and in some industries scarcely that. It’s a wonder that it can work at all.
But couple that with a parasitic investment community that insists on having dessert first, front loading flows to their NPV – at a discount – outsourcing not just production, but capital; and see what you get. A deeply indebted cash-flow business that is wholly dependent on a strong dollar. Recapitalizing the banking system only makes it still more tenuous. On all counts. The systemically important part is the part that creates the wealth. Not the part that intermediates those flows.
So Anon of 6:46, agreed indeed. Living simply and working less for The Man is better, and I enjoyed my life better when I was doing that. Then the bubble drove up all my costs while The Boss forced hours into the fewest possible workers, ostensibly for benefit issues. And here I am, trying to get back to the Garden. We should all work less and in accordance consume less.
Pudentilla and MarcoPolo, agreed on all points. We have lost sight of the importance of real production in this country—and not by accident. Real production was heavily unionized, and even where not indexed by the set of union-defined scales and benefits. Then A Certain Political Party and many likeminded set out to destroy labor, for social reasons and to get more of the beans for themselves. "Unions bad; off-shoring good; trickle down will float all boats (you stupid marks)." Turns out those universal masters weren't all that smart, merely clever.
The establishment is trying to convince us that tomorrow will look just like yesterday.
I want to believe, I want to believe……
nawwwww.
Every study ever done says the majority of companies are disappointed with the results. It seems to represent hope over experience.
Is this a political comment?
In all of these studies you are reading on the issue, are they also stating "empirically" that not only is outsourcing inefficient, but "we have found no progress over the past decade at reducing these inefficiencies".
Well I can speak from personal experience about what is obviously only one type of outsourcing.
I've been involved, at the nuts and bolts level, of implementing an IT offshoring arrangement at one major company (over about 8 years) and then at another company with an already established offshoring approach (for the last 5 years).
I would say the 25% (IT (software) productivity is extremely hard to measure unfortunately) sounds feasible to me after a decade of working on it. I honestly think we were truly losing money in the beginning.
Management saw a too good to be reality, they could buy 5+ hours of developer time in India for what 1 hour costs here. Surely inexperience, cultural, language, and timing issues couldn't eat up that big of margin. But honestly I think it did. In the early projects we'd go through 3 companies before we got anything vaguely acceptable as far as productivity. Even now what you see is the onshore personnel have to be extremely good and work extremely long hours to make the offshore ones productive.
And the thing is it's not a question of skills. It's a misunderstanding of the process. Everyone wants to believe that requirements and specs can produce good software from coders without them having any understanding of the business, the existing system, the upstream and downstream systems, etc… but in the real world it doesn't work that way.
Coding is easy, knowing what to code and what not to and why is the hard part. And offshore personnel never have the opportunity to learn those parts, so the efficiencies are effectively capped. just my opinion.
Sometimes you outsource tasks simply because they are not part of your core competency. For example, maybe you have relatively little hardware engineering capabilities/needs. Rather than try to do the hardware engineering in house (with a staff that is less than critical mass for that engineering function), you simply outsource the function to a company that specializes in that discipline. It is obviously the right thing to do some of the time.
A political/economic reason for outsourcing (if you are a multinational), is that you want to penetrate the market in a particular country and that country demands that you throw a bone its way to do business there. That bone might be an engineering facility located in the desired market or maybe outsourcing some production to that same country. That is the way the real world works.
Outsourcing is not always the way to go but sometimes it makes excellent economic sense.
While Taleb makes good points our problems run deeper then a financial fix held up by political corruption. World growth dynamic is related to cheap oil which has been responsible for driving the standard of living and with the picking of the low hanging fruit coming to an end our economic system based not on some grand theory of economics but cheap energy can no longer produce the sustained gains in GDP necessary to satisfy the computer models.
Who would know how to do this?
Nobody.
After PPIP fails (because it's a stupid idea), and Obama tells the the Treasury to get out of the way, the FDIC can go in and start making the banks charge the crap off, letting the appropriate heads role. Then a real market for the crap may begin to function.
We need to burn off the bad debt. That requires recognizing it. Cheating the holders of good debt of their value through massive equity dilution–not a good idea.
Taleb and Spitznagel's simplistic solution is fraught with moral hazard. Our defunct social contract is at the root of our problems, in my view.
No more shell games. Let's get back to values like honesty, integrity, honoring contracts, and making failure possible.
No more "too big too fail."
"Instead of sending hate mail to near-insolvent homeowners, banks should reach out to borrowers and offer lower interest payments in exchange for equity."
Would someone please explain what this means? Who gets equity, in what, in exchange for what?
Anon of 10:16 AM,
The studies I have seen (I will confess I have seen a fair number) were surveys of the parties that did the outsourcing (ie, the corp clients) by folks like Deloitte Touche, who presumably have no axe to grind and it anything might benefit around the margin (any byg corporate change tends to create extra work for advisors of various sorts). So I am in no way, shape, or form referring to studies done by groups that might be engaged in advocacy (ie, have tied to or are backed by pro labor groups). And per below, lawyers, who are starting to feel the impact, are having it happen as a senior vs. junior phenomenon, and with senior people not affected (yet, much if at all), the people with the money and connections don't feel compelled to take action, particularly, form an industry association and get a message out (the usual way things like that seem to be handled now).
One reason, in fact, for the surprising lack of publicity around these disappointing results is that they tend to affect white collar workers, who are not organized (the notable exception, AMA, accounting industry, lawyers, which are de facto craft unions, whose restricted entry serves to keep labor costs higher w/o formal wage setting goals of industrial unions).
Steve Koch,
Fair point, but most of the outsourcing I have seen (ie, the kind that will arguable impact costs and please Wall Street) is pretty large scale and does involve core competencies. For instance, I read on Slashdot that outsourcing is so endemic in IT that it is well nigh impossible for young IT professionals to find a career path, that is, go to a big or biggish company and get meaty experience and develop. Now that might be overstated, there are lots of places too small to outsource, but any small firm is less stable than a big company, both from an employment and quality of work standpoint. It is often lamented at Slashdot that we are hollowing out our IT industry, once an area of US leadership.
You see the same trend to a lesser degree in the law. Large firms are demanded by some clients to outource research to India. Again, this is a way to develop juniors, Again, it is to the detriment of core competencies in the profession.
We outsource becuase it's what we teach our MBAs. Outsource! China! India! >> degree
Well said, Walter. IT software productivity is notoriously non-commoditizable, as the title of the classic "Mythical Man-Month" (sitting on my shelf) by Fred Brooks attests. Fred found back in the '70s that it was not uncommon for one developer in a group to be ten times as productive as another, and this is still true today. This also explains why things like pair programming (having two programmers program together at one computer) can work, even though it would be anathema to a manager measuring man-months.
@Yves Smith said: "One reason, in fact, for the surprising lack of publicity around these disappointing results is that they tend to affect white collar workers, who are not organized…"
I wonder if it's not more complex than that. IN 1932 Reinhold Niebur wrote:
The white collar worker may not own any property and may therefore logically belong to the proletariat, but the dictum of Boudin and others that salaried workers "are in reality just as much a part of the proletariat as the merest day laborer" fails to take important psychological factors into consideration. If we may regard Germany, where all the social and political forces of modern civilisation have reached their most advanced form, as a criterion, none of the disinherited middle classes express themselves politically in proletarian terms. On the contrary they turn to fascism, which combines enough radicalism, to give the poorer middle classes some hope of better things to come, with the political strategy of anti-Marxism and nationalism, by which it gains the support of the economic overlords, who are afraid of the rising tide of labor. That the middle classes can be drawn into a party in which the wealthiest and the poorest ostensibly make common cause, is the measure of their (the National Socialists') political intelligence. Whatever may be the logic of their position in economic terms, they (the poorer middle classes) would rather express their resentments in a nationalistic spirit, and in minimum demands for the elimination of financial abuses, than in thoroughgoing economic changes. They will never be reduced to proletarian terms politically (even though they are reduced to those terms economically) until they have lost their culutral as well as their economic inheritance. Unlike the proletarian, they do not stand outside, but thoroughly inside, the national culutre.
~
I believe that even though Niebuhr's predictions proved correct for 1930s Germany, they nevertheless did not prove correct for 1930s United States. In the U.S., the factory worker, the small independent farmer and the poorer white-collar worker did did find solidarity. "Thoroughgoing economic changes" were demanded and became law. Later Niebuhr was to hail this as "experience triumphs over dogma," an American labor movement that "was a pragmatic movement, born of the necessity of setting organized power against organized power in a technical society."
"For the life of me, I never understood the vogue for outsourcing and offshoring"
The people who actually set up factories offshore argue that it is necessary for the companies, so that the companies remain competitive and remain in business. Which is true, at least for manufacturing.
Outsourcing white collar jobs means all-in cost savings of 15% to 20% and you call that a mere savings? Why is it mere?
Outsourcing and offshoring are reasonable responses by companies to the business environment. The real culprit is the economic policy of the US, or lack thereof, that creates a business environment relatively hostile to domestic job creation, capital formation, and productive investment.
This is an interesting topic, but I must say I think Anonymous Jones is pretty much on the money for me.
Anon of 4:11 and Millers,
I did not tease out the manufactuing part of the argument, but I have been told by executives by US manufacturers that moved manufacturing overseas such as Ethan Allen, that there was NO economic benefit for doing so (transport costs back and forth, have to finance inventory). An extended supply chain with big overseas pieces makes it hard to do just in time type strategies, which can be very cost efficient. You need lots of buffers.
And the people who told me this were NOT the manufacturing types, who might have personal axes to grind, but top staff, like the general counsel . This was what Wall Street wanted to hear, so the company complied.
I have been told more broadly that there is no reason for the US to have ceded anywhere as much manufacturing as we have. Reject rates from China even now are still very high, Conversely, in many industries, more flexible, shorter-run, more customized end user focuses approaches could hav e been very competitive in a lot of industries. But manufacturing is dirty old tech, no one values it any more.
Similarly, the loss of the domestic shoe industry (I have read several studies supporting this analysis) was due to the liquidation of Interco, in the aftermath of the last LBO boom (late 1980s). Interco had a funding source for a Ch. 11 lined up, but it fell through at the very last minute and the BK morphed into a Chapter 7. Pretty much everyone who knows that industry says the shoe companies in Interco, which accounted for the vast majority of the US industry, were viable even in the face of overseas competition.
As for the 15-20% savings, did you read what I wrote? I said relative to risks and rigidities. Look at the simply massive coordination costs, the labor savings (large) relative to overall savings (much much smaller). More coordination means more opportunities for miscommunication and screw-up, PARTICULARLY when you go across national borders, even more so when people whose native languages are different deal with each other. I have done a lot of international work, and this is a bigger issue than most people in the US realize.
And if 15-20% saviings was what the plan said, what they realized is almost certain to be less. There are reasonable odds there were no net savings, even a cost increas which would be massaged to hide that the effort was a failure.
And if the deal involves outside parties, the rigidities are HUGE, and if you ask for anything not in the contract, you get hit with very costly change orders. And if they do not perform up to your quality requirements, your only way out is typically to exit the agreement, which again is very costly and disruptive. What good are theoretical savings if you annoy, worse lose, customers?
My brother happens to be in this business, so I am not speaking from a theoretical vantage here.
So we convert debt to equity and all of our problems are solved. And how exactly does this bring back all the millions of jobs that were lost? Oh it doesn't. So what does this solve actually? Companies will have better looking balance sheets for a bit but only for a bit. They won't be able to sell much to un- and under- employed or to those employed who have been scared into saving and paying down debt. Their demise will only be delayed not averted.
And what is this hoohaw about banks lower rates in exchange for equity? If mortgagees are upside down in their loans, there is no equity.
The key paragraph for me though is this one:
"We believe that stimulus packages, in all their forms, make the same mistakes that got us here. They will lead to extreme overshooting or extreme undershooting. They lead to more borrowing, by socialising private debt. But running a government deficit is dangerous, as it is vulnerable to errors in projections of economic growth. These errors will be larger in the future, so central bank money creation will lead not to inflation but to hyper-inflation"
This exposes the underlying agenda of Taleb et al. They are inflationists. Can't have that debt feeding into inflation, no, no, no. But where is this inflation they are talking about? We have mountains of debt, debt as far as the eye can see, debt out the wazoo, so where is the inflation?
Now someday we may face inflationary problems but the truth is now the danger is deflation, and for this we have real evidence in the decrease of aggregate demand, which would only be exacerbated by the anti-stimulus demand destruction Taleb et al are advocating. I mean just look how weak their argument is: overshooting or undershooting or missing projections, like that has never happened.
And just to be clear I favor some debt to equity conversion but it is just one piece in an overall redistribution of debt to all the participants in the financial system.
I wrote a comment somewhere saying that certain debt (especially debt issued by publicly-traded financial bodies) should have 'reverse-convertible' characteristics.
i.e. a debt to equity conversion process.
This would link the debt, CDS and equity option markets.
Banks issuing debt would automatically be buying out-of-the-money put options which would raise the effective interest rate acting as an insurance premium to exchange (principal) payments for equity should their equity sell off.
Yves,
OK, let's restrict the outsourcing discussion to that of outsourcing core competencies to other countries. I'm not yet convinced (though I am willing to be convinced) that businesses are usually losing money when they do this. If they were, I think it would be much less popular. American managers tend to be short sighted and don't tend to look much beyond this quarter's results.
BTW, what percentage of the time do you think this kind of outsourcing is losing money?
I think outsourcing our jobs to other countries is a big problem for the USA but it is inevitable if you embrace globalization. I don't know what the solution is but we have to do something about it. Maybe we tell them we buy as much from them as they buy from us.
For the last 20 years of my career, I worked primarily with non-Americans working in other countries (i.e. not the USA). For a multinational, it really helps to have local people (in China or Brazil or Mexico or Russia or Norway or wherever) intimately involved in the engineering and/or manufacturing process.
My experience is that if you want to do a lot of business in another country, they tend to demand that you give them some manufacturing or engineering jobs.
There has been huge progress in communication networks and translation facilities. The world is getting smaller. In software engineering, the reason so many jobs have been outsourced is that there are standard languages across the world. .Net or Java is the same in China or India or Russia as it is here.
In my corporation (which is an elite engineering company), HR wanted us to hire as many female engineers as we could. We had a difficult time finding/hiring elite engineers who were American women so we started hiring female engineers from China in large numbers (and we hired a bunch of male Chinese engineers, also). After a while corporate figured out that it would be cheaper to hire/employ Chinese engineers in China rather than America and opened an engineering center in China.
Hugh:
I think the idea is to change the loan so that the lender gets a percentage of the house (i.e. it is not just collateral). If the lender gives the borrower a lower monthly payment in return for a percentage of the house then the borrower will be more able to make his monthly payments while the lender has a chance at a future profit. At least that is what I have been thinking makes sense.
The borrower still gets to build some equity and not screw up his credit while the lender has the opportunity to make money on the chunk of house he bought.
Ahem. A significant other problem has been that assets have inflated far beyond wages for about the past 30 years.
So, if we're limited to either deflating debt, or inflating assets, without taking into consideration that we've alredy significantly inflated assets while wages have stagnated, misses a big part of the picture.
We face deflation of demand in large part because people have become aware that their wages alone can't support the consumption they were engaging in. They came to that realization when they found that inflated assets can evaporate, while debts and future needs are less flexible.
Steve,
You seem to have misconstrued quite a lot of what I was saying.
First, while I did suggest in the case of modest on paper cost savings, they could turn out to be zero, the much bigger point in on a risk adjusted basis (risk to the business from rigidity, lack of responsiveness, longer lead times, more communication nodes leading to more points of failure) a lot of outsourcing was dubious.
If you were doing a proper corporate finance analysis, you'd attach a much higher discount rate to the profits in the outsourced model because much greater risks are involved, From what I can tell,that was very rarely if ever done. Everyone looked at it on a pure cash basis.
As for the US manufacturing examples, I am not talking about producing locally to serve local markets, I am talking about manufacturers with significant US customer bases choosing to serve them from afar, typically China or Southeast Asia. You bring up a conpletely different scenario.
As for the international bit, I am well aware that computer languages are standardized. That is not the point I was making. The needs of the US customer need to be converyed to the programmers overseas. I do a lot of transactional work and I see miscommunications among people who use the same words but actually mean something different all the time, And that happens much more often when you go cross border or cross culture (as in accountants versus salespeople).
More people in the communications chain increases the opportunity for screw up. And there are cost-effective, rapid development approaches, like Extreme Programming, that are not well suited to a long distance model.
There are considerable advantages to face to face contact and ease of access. Those get lost in these computations.
So Walter and Steve Koch, here is a single word that links our signficantly different perspectives: expertise. Walter, I appreciate hearing your perspective on this from one who has done it, that there are evidently hard caps on gains from outsourcing. As I understand you, moreover these are significantly because distant and culturally differnt producers cannot duplicate local expertise but only compensate for it with significantly longer hours.
And Steve Koch, the issue of core competencies to me is often misconstrued because it undervalues expertise. Your basic point is perfectly sound, that there are instances where outsourcing is a good econmic decision for a firm. For instance, if one can buy a product or service off-the-shelf where robustness is important but expertise is lower, it may even be the best decision to turn to an outside vendor rather than duplicate a product in-house at no real advantage. Getting this call right is easily done wrong, but presumably this is what upper management is payed to get right.
The issue of what is a core competency is made from management's perspective on costs. Even things which are _not_ core presumably can matter greatly to the success of the outcome if they are only reproducable externally with a loss of quality. You don't have to 'build it/run it yourself' but the result has to be as sound as if you did at least. And if outsourcing lowers quality or impedes your overall efficiency, the cost saving is offset by a productivity/quality degredation. I suspect that you're familiar with this in the course of setting up shops offshore as you describe. [cont.]
Expertise is generally undervalued. Knowning how your manufacturing works and why isn't just a question of coding or inventory management. Knowing what works with your customer base requires a certain degree of engagement with those customers which those distant in space, time, and culture will lack even if equally competent at the 'core' operations involved.
Expertise is undervalued for two (at least two) reasons. First, it is bleedin' difficult to quantify, so it's hard to explain to your boss. And expertise is highly specific to experience. Even an outstanding manager may have personal expertise only in a small area of a firm's overall operation and customer interaction; certainly so in outfits large enough to offshore chunks of themselves. You know expertise is gone only after you remove it from your system. And if you as a manager just bet your career on a ten-year redesign including outsourcing it will be extra hard to acknowledge that little was gained, and highly unlikely to be acknowledged that something was lost. This is why it takes _outside assessment_ such as Yves cites for firms to establish that their ambitions upper management has decimated their expertise base, quite possibly unknowingly.
Second, expertise is undervalued because while _existing_ expertise benefits the organization as a whole, because it is pre-existing it does nothing to help a manager make their bonus. The organization already has it so one gains no personal credit or extra remuneration for maintaining it, in the US model anyway. (In the Mittelstand model this is quite different.) In order to justify getting a bonus, one has to show that one did something better, or failing that something new that wasn't (obviously) worse. And showing that the Firm saves $X by outsourcing justifies a bonus number. Showing the the Firm eliminates permenantly X while generating 'the same' output definitely justifies a fat number. So the bias is to look at the number not at the unquantified in-house expertise. This is part of the American Stupid corporate model.
The bonus system, to me, is a pernicious destroyer of organizations because it teaches everyone eligible to defect from a common goal but maximize their personal position. The results are certain to be a net degradation of capacity for the organiztion as a whole. Yes, bonuses may motivate extra effort, and some forms of productivity may increase. But they are toxic overall.
Yves: " . . . [I]n many industries, more flexible, shorter-run, more customized end user focuses approaches could have been very competitive in a lot of industries." This is the direct example of what expertise and local interaction gain you. You have the potential to eliminate miscommunication and refine implementation, in as near to real time as one can do for large organizations. As Walter says, it takes exceptionally competent people working significantly longer to compensate for having to work at a distance, when it is possible to so compensate at all.
Richard Kline – was about to pile in on the expertise point but you have beaten me to it comprehensively.
Perhaps the round up is that outsourcing ignores externalities (supply chain risk, ease of communication)? The expertise of company employees isn't on a balance sheet either but it gets valued at zero too.
So Richard Smith, I don't know that (smart) companies value their expertise, both in employees and via intrenal systems, at zero, but the certainly don't value them at par. (Except perhaps in small business and partnerships where the importance of the others' contributions is before ones eyes.) And I would suggest that the further the evaluator is from the evaluee, the greater the deflator in their estimation of the worth of the expertise involved.
If you have your own take, it's not to late to put in your observations RS. Folks do go back and follow comments for several days in my experience, especially in long threads with much meat such as this one.
Mmph yes, you are right, "zero" is hype, at least in general.
There can be specific cases where corporate reluctance to admit that the expert peons do have a negotiating point leads to ridiculous levels of risk.
One in particular is burned into my brain just now.
The client was a big outsourcing fan and had the mindset where techies are just coolies and price is everything. They jibbed at the cost of employing the staff who looked after the system. But their outsourcing suppliers were (wisely) not too keen on contracting to support the system if there were no experts around to consult.
Client took the system in-house, hired a few of the experts as contractors, in a desultory manner, then failed to hang on to them: there was a corporate rule that contract durations were a maximum of two years. Works if your hired bodies are interchangeable. They followed that rule in a case where the hired body mattered.
The upshot was that a business critical custody and dealing system that needs a chunky expert staff to look after it (say 2 millon lines of quirky legacy code) is now in the hands of two terrified newbies who have a subset of the technical skills required, none of the business knowledge, and a fragmentary knowledge of the way the system actually works. The first regulatory change or, worse, operational hiccup that occurs, and my erstwhile client is in deep trouble. He doesn't even have the knowledge in house to move the data off onto some more maintainable system.
Plenty of bungling there of all sorts, but the failure to value the IP properly is somewhere near the middle of it.
Your point on the remoteness of the decision makers from the coalface that would give them the needed info: smack on. I would add that there is a sort of cultural bias in that type of manager: since they don't know the detail, and have learned to prosper, in a career sense, without knowing the detail, they systematically undervalue it.
Yves,
OK, let's further restrict the discussion of outsourcing to US companies that sell only in the USA. The companies that I worked for (Intel and Schlumberger) compete very successfully internationally so I have less experience in the constrained model you propose (but a lot of experience at what you have to do to compete successfully world wide).
I'm not sure why you think foreign vendors are necessarily going to be more rigid, less responsive, slower to respond, less dependable and more difficult to communicate with. It depends on the vendor. You are going to tend to outsource your work to somebody who is world class. Companies get to world class status by being flexible, responsive, dependable, and communicating well.
We closed an engineering center in the USA because the work had been outsourced but a big part of the problem was that particular engineering center performed poorly for several years. It was given every chance to succeed but they still failed.
When you outsource software on a big scale, the remote company most likely will deploy people to your facility to simplify the communications issues.
There is no doubt that it is easier to talk to the guy in the office next door than to talk to a guy in Cambridge, England (for example) but the reality is that sometimes the expertise you need is located in Cambridge, England (not next door).
Sometimes the best people for the job are not going to be Americans. In recent years I worked quite a bit with particle filters. I was fortunate enough to work on this topic with some brilliant math Phds in our Cambridge, England Research Center. We opened that Research Center in Cambridge because we wanted to hire super sharp grads of Cambridge University. It worked out great.
Outsourcing does not always makes sense. It is hollowing out America. But part of the problem is that we need to compete better.
Richard & Richard,
The experience and IP issues are very interesting. Regarding IP, I worked for the manager of a large engineering center who did not know the difference between reverse engineering and straight theft of intellectual property.
Corporate IP is valuable and there are lots of people who want to steal it. Outsourcing increases the opportunities for IP theft but hiring foreign engineers does the same thing. For that matter, there are probably quite a few American engineers who can be bribed to sell corporate IP.
Experience is under valued. Corporations tend to be weak in their ability to institutionalize experience. Outsourcing can actually force a company to institutionalize experience (i.e. to enable the external vendor to work effectively).
Steve,
I think you may be talking about a different type of thing from Yves' hobby horse. I agree very much that the search for talent can lead to some pretty strange geographical tie-ups that actually turn out quite well. A tiny scale illustration: I worked in a London IT shop that had a first rate guy doing business analysis for UK markets – out of New Zealand! Once you got used to only being able to confer live with him for an hour or so at the beginning of the day, just before he went to bed, it all worked fine – his specs were well up to scratch, so that he really could hand the spec over to developers to code, and the final product was consistently what you wanted and on time. On a larger scale in another different life I remember scouring the face of the planet for Surface Acoustic Wave engineers, who seem to be as rare as unicorns. I think we got a bunch of Canadians, in the end, if memory serves. For certain, we looked there, meaning to take them on if we found them. I think when talent is the driver, and people are trying to push the business on, then all sorts of expedients can be made to work. Location and time zone aren't such a big deal. You take the talent where you can find it. Same for engineering suppliers as for hires.
By contrast, with these cost-driven outsourcing deals, where you are taking some legacy software or other and dumping it overseas, and claiming you've saved some maintenance costs, I am pretty firmly with Yves: the interfaces seem to work pretty badly in practice, and the cost reductions seem modest for the disruption and risk, or nonexistent. It's a talent thing: I really wouldn't hand over some of the specs that I have sneered at in banks recently, to some polite guy overseas who knew nothing about the business and systems context. It doesn't matter if your silk purse makers are 20% of the cost of the locals: if you hand over sow's ears, nothing good is going to happen. Yet that seems to be the model. It happens all the time. It's a sort of travesty of engineering.
Management disillusionment with the costs and risks of large scale software development may be a driver driver: through the 80s and 90s companies wasted prodigious sums of money on ambitious software projects that failed miserably to deliver. The ones that made it are a random hotchpotch of different technologies of varying vintage.
So now there is a trend the other way: instead of overspend, the emphasis goes into cost cutting, and investment is pared back (way below the prudent minimum IMO). And we get a really clunky process, with murderous change control provisions in the contracts. Such fun to manage.
I'd better stop ranting…