Limits Of Incentives

Limits of Incentives and the changing paradigms in a downturn By Procyon Mukherjee

Incentives seem to be the magic word driving change in the modern world; they act to steer choices in a certain direction, whether it be government, the private sector or even in education. Incentives act to influence the pay off structure of the utility function; it helps to create value for both the transacting entities. Between the two alternatives, incentives serve to offer a better choice in which there is gain from both sides. The transaction can only happen if both the parties find the course of action a better path to take in presence of incentives to do so. In absence of incentives the outcomes could be quite different and less efficient. The origin of incentives can be traced back to the government actions to generate employment or generate money for the treasury. In absence of incentives both these outcomes proved to be much less impactful. Whether it is providing tax breaks for setting up new industries or providing retention incentives, the government acted to generate employment in a certain region and continued to incentivize such actions to retain the benefits on an on-going basis. This extended to providing incentives on investments or on growth (sales volumes) or on cost (10% cost reduction). The results of these programs vindicate the fact that incentives tend to channelize actions towards maximizing the objective function or the pay off function to the mutual benefit of both government and the party involved. The act of incentivizing people to deliver superior results took a much forceful connotation in the 1990s, when there was an increased focus put on creation of shareholder value (some trace this back to the speech by Jack Welch, but actually he did not mean to say this). Management started to believe that if a strong top down incentive structure could be put it could drive behavior towards generating greater value for the shareholder. The shareholders believed it too and there were strong evidence in favor of such a denouement particularly in the boom phase of the business cycles. However such incentivization failed to return a similar end-result in the phase of the downturn. But since the world had seen very mild downturns it never had been an area of study, whether one could continue to incentivize people using the same yardsticks in a downturn as in an upturn. Incentives designed to drive behavior are being realigned after the current debacle in the financial sector. The U turn is equally significant. The very recent example is the putting of 90% tax on Wall Street bonuses on executives of firms bailed out by the tax payers. Tax payers as owners or the new Principal are putting in controls to see that large part of the share of the incentives is routed back to them. The question is why this was missed in the upturn where the Principal parted with a large sum without insuring the risk? Is the recent action by the ‘new Principals’ going to trigger any different kind of behavior? Is the latest action going to bring in more risk averseness that would not improve performance at all, which is the very core strategy of incentive design?

I would try to deal with the issue of incentives in a downturn. When excessive risk taking was rewarded during the upturn without imposing limits (Akerlof and Shiller in their book Animal Spirits bring out this aspect quite succinctly), excessive risk averseness cannot improve performance and if incentives or avoidance of incentives spur such type of behavior, the downturn only would be prolonged. Downturn cannot be whisked away as an external event. Managers have to deal with the issue of a deep slump in the market, which are partially their own creation; no one can shirk away from the responsibility. It was Jack Welch who last week pointed out that it is not the GDP that influences company performance but it is the other way around. It is the performance of companies that influences the GDP. So it is the end of pointing and giving excuses. In the same discussion he pointed out that creating share holder value is no strategy, it is an end result (FT 12th March ’09). So the current style of leadership has to change and the way to change is to design incentives so that the right behavior is demonstrated that would lead to the right actions. Incentives in an upturn were designed to spur a positive movement, managers, at least some of them misused them and took excessive risk, without informing the Principal about the implications in the longer term. This I would term as a clear case of failure of incentives. The most recent case of incentive failure is so much in evidence in the crisis of banks. The bank CEOs were given Return on Equity as the key performance criteria for their incentives. The CEOs therefore indulged in excessive risk taking that would lead to increase in net income (the numerator), but simultaneously ensuring that the Shareholder’s equity (denominator) did not increase. The net income for banks can only be increased by higher lending (and riskier lending); it would automatically mean raising more capital either in form of debt or in form of equity. The Bank CEOs did not raise capital through the equity route, thus when they moved to higher gearing ratios or higher leveraging there debt equity ratios exploded. When the asset prices collapsed, their liabilities remained while the assets shrunk in size, thus their net worth shrank. If their incentives were not based on Return on Equity as set by the principal (the principals wanted to reap benefits of all the profits without making the pool larger for sharing the profits) perhaps the banking crisis could have been avoided. However the increase of assets that turned toxic was a wrong doing that entirely cannot be attributed to the incentives or the failure of incentives. It was partially due to the absence of proper stress testing systems and procedures in the banks and the proper understanding of some of the risky instruments was lagging in some cases. However the initial motivation to move towards this higher risk portfolio stemmed from the misalignment of the incentives with the risks that the CEOs were taking. There is another behavioral aspect that one would notice in an upturn. The growth of markets and explosion of commodity prices gave enormous profit opportunities to most companies. The incentives merely based on the profit metric were unable to segregate the market effects. But the managers being motivated by the incentives overlooked the fact that the profits were a proxy for the market sentiments and were they to return to its normal levels the profits would have melted. But the self-deification by managers in an

upturn when the prices moved up exponentially point to the misalignment of incentives with actual managerial performance to drive share holder value for the long term. The incentives missed to capture the regression with the market. For example if a company improved its top line due to sheer improvement of market fundamentals that came from the growing economy, then that should have a regression with the company’s growth and incentives should factor this regression. This is true for pricing as well. This would tend to constantly keep our focus on internal actions which are specific and measurable that drives results and therefore would continue to play a leading role in spite of markets performing differently when the business cycles change. This would actually provide as insurance during the downturn that many managers would fail to recognize. Stimulating the right behavior should be the motive of incentives. Channelizing actions that would actually increase risk instead of reducing it is a classic example how incentives failed to stimulate the right behavior. The agent in this case was also at fault to have missed the over-sight processes to see how it was itself bringing its own peril. However thankfully there are better examples already available that some companies have beaten the market downturn and have returned better results. The example is Nestle or Mitsubishi conglomerate. Such companies however are rare who could lead and be outliers. Nestle kept a careful eye on the range of products it launched and the target markets where it launched them. While it had products that were of high value that the rich only could afford, it had similarly products that could be targeted to low income families as well. In a recession these two types of products targeted at two different segments helped to fetch them superior returns compared to the competition. Mitsubishi on the other hand as a conglomerate worked on the principle of saving during the upturn to be able to invest and buy businesses in the downturn, an unusual strategy that made them aware of the limits of markets in the boom time and when to withdraw in these extreme times of exuberance as the prices could be irrationally beaten up without any logical reason. There are examples in the banking community itself (Hudson City Bank Corp. where net income rose 50% last year as per Wall Street Journal 23rd March), where in there are some of the smaller banks in Wall Street itself that did much better even during the Banking crisis. Do the same incentives that generated so much of positive returns (till they turned toxic as well) apply when it comes to the downturn? Is there a need to change the very structure of incentives to avoid excessive risk taking? Can behaviors be actually changed through the incentives? There is an asymmetry that one would realize while looking at the same incentives during an upturn as in a downturn. The whole designing of incentives need to be carefully reviewed if this asymmetry has to be resolved. If managers are rewarded when it is the upturn of the business cycle that contributed to wealth creation because of a multiplicity of factors both external and

internal, they should not be rewarded again when it is the downturn by the same logic, the downturn itself cannot be sighted for the lack of performance. But this puzzle of incentives could be tackled by segregating the external factors from the internal ones, but it would call for alignment of objectives between the principal and the agent. I would like to go back to the Jack Welch example, when he sighted the example of GDP. Text books would define GDP to be the cumulative income of all the individuals and companies and government in an economy. Although there is an intertwined relationship between the income of any company and the income of the whole economy, but one cannot simply use the logic to suit an end. The fallacy of the logic can be seen if we try to see what actually happens when the income grows. The income of a company can grow when the sales growth takes place because of an innovation or the cost comes down because of a productivity increase. These are the basic factors for income to grow for an economy as well. However there are economic conditions either created by the government or by the market or by both which aid this process of income growth. By pointing at the conditions alone the managers cannot shy away from the responsibility that they could not raise productivity standards or could not create innovation, which are the basic functions of management. However for management to do this, there are incentives needed to drive a particular type of behavior. The principal or the owner for whom the agent or the management works create these incentives to serve the purpose of ‘motivating’ the change process to reach the desired objectives. The history of incentives however is rather short, at least the monetary part.

I would like to take examples of incentives in the area of Electricity tariffs that drove a different kind of behavior amongst the consumers and thus helped to create a win-win condition for the utility companies as well. We all know that electricity consumption is never steady, it has its peaks and valleys in the different parts of the day, peaking during the evening hours. While this is the case for the urban consumers, for the offices and industries the consumption pattern could be quite different. This consumption pattern would mean that while one class of consumers are not using electricity, the no-load losses increase for the utility, while when another class has the increased need, servicing that need also becomes difficult because it would mean creating a higher supply capability only for a short duration. Added to this is the problem of cost of electricity itself, which is different for the different classes, as the transmission and distribution losses are different for the HT consumers (Industrial) as compared to the LT consumers (domestic). This asymmetry riddle could be solved with the design of incentives for one class of consumers, because otherwise it would have meant one class subsidizing for the other, which would not have been a good practice. The solution was that the industrial consumers were given something called the Time of Day Tariff (TOD) which meant that they had different rates for the different times of the day, the highest tariff was accorded for the peak hours (evening) and the lowest for the off peak hours (late night after 10 p.m.). What this meant was that an industrial consumer could gain by maximizing its use

of electricity during the off peak hours and could lose if it wanted to use it in the peak hours, thus the utility function could be maximized for both the parties, the HT consumers and the Utility itself. This kind of incentives drove the industrial consumers to design their processes in such a way that they could get the maximum benefit by using as little electricity as possible in the peak hours. The shift pattern of consumption got completely revised in some industries to get mileage from this. Thus more power was available in the peak hours to be distributed to the domestic consumers. This incentive program of TOD Tariff was a significant change in the history of incentives, because it was for the first time that the importance of collective benefit was understood in the realm of incentives. The same incentive problem when looked at from the domestic consumer’s view point who is living as a tenant in a house whose electricity bill is paid by the owner of the house, does not drive a savings motive. This is the example of moral hazard, because his honest intention to save does not give him any pay off and his behavior of wasting energy would also not lead to any penalty. Thus an incentive program that does not take care of the moral hazard problem is bound to fail; the best example in our current times is the issue of toxic assets and their disposal. The government through the designing of incentives have failed to realize that the public private partnership cannot succeed unless the issue of moral hazard is solved; increasing pay offs to the private sector for the gains and extending the ownership for the bulk of the failures and the pay offs thereby to the government would not incentivize taking cautious actions that would stop creation of such toxic assets. When we talk of incentives we need to keep in mind that to stymie any effort that leads to counter-productive denouements, should be meted out with disincentives or punishments as well. To stop moral hazard the only way out is to provide for punitive measures that would restrict people from showing that type of behavior. If the tenant is penalized for wasting electricity that is paid for by the owner of the house, surely the tenant would behave differently. These examples bring us to the central theme of any incentive program; it must be able to drive a certain kind of behavior that gives benefits to all constituencies. Giving benefits to some while delivering little to others, does not lead to sustenance of the program. Also there must be punishments instituted to ensure that one would not go beyond the limits to bring peril for the long term. This problem can be seen in the area of incentives between a customer and supplier in sharing profit. When one class excessively tries to derive all the benefits by virtue of its superior positioning or power, the incentives for one turns out to be the disincentive for the other. This is also a case of failure of incentives. For a particular duration it works fine but the one from whom the incentives are gradually withdrawn would be waiting for the opportunity to strike back as well. Many of the contracts signed between two parties are found to have one sided incentives. These contracts do not generate long term value for both the parties.

Let us go back to the example of the banks and the incentives. The failure of the incentives was on many counts, I would like to point out to some of them. 1. Return on Equity as a metric sparked off a deluge of actions that exploited a given market situation that had a relatively low risk free rate and people were looking for an opportunity to invest in the right engines of growth. The rise in money supply sparked off an asset bubble that the banks helped to create but not stymie it as the incentives were not in favor. The market amongst many things caught on the bandwagon that more demand for assets would essentially mean more asset prices, regardless of the economy’s ability to sustain such demand pattern. The overall gain for the society from these incentives were such that each gained from the run, and if the gains of one were to be reversed it would have brought peril to all others. This clearly is a seed for the eventual collapse. 2. The cap was missing: The incentives never had a ceiling, thus there was no end to one’s rewards, which would mean that one could increase the stakes to any extent. The principal was initially rewarded handsomely from this run, so there was no need felt to cap such incentives with a ceiling, because they probably felt that it would be bringing in a ceiling for their earnings as well. 3. Incentives were unidirectional and did not look at all the constituencies and the various elements of risks and their impacts. There was no parameter that could weigh this aspect for the medium term or longer term. Since Bank balance sheets could hide some of these portfolios and there is actually evidence that off-balance sheet transactions stymied the effects of some of these excessive risk taking overtures, there was no way that anyone could clearly discern the true performance after factoring the risk. By spreading the risks, the banks thus created a common risk pool for risk, thus alienating themselves from the eventuality as their ultimate failure could never be traced to one single entity. 4. The regression was missing with the market factors when incentives were designed, there was a hallow effect that had to be discounted, but which no one did. When asset prices assumed the most outrageous levels people believed that it was never the peak (Some Banks predicted oil to reach $200 a barrel when it was hovering around unprecedented $145 per barrel) 5. The incentives used much higher discounting rates, meaning that pay offs were higher in the shorter term than in the longer term. This last point needs to be elaborated. The incentive design is based on short term results, which is very important. But the pay-offs could be staggered thus creating the need to see that the results are sustainable. This wasn’t the case because the banking community assumed much higher value to the discounting factor for money. Money now was made to look far more precious than money two years from now, where as the long term interest rates were actually hovering around 2.5% to 3.5%, not a very great number. This looks rather odd and highly unexplained human behavior. However the stock options as a vehicle to drive behavior helped to support this long term aspect, but only to a limited extent. The stock options with the right kind of strike prices

(presumably using the Black Scholes model) helped to do two things, firstly it aligned the agents motives with the motives of the principal and secondly it helped to create a long term view of results and pay offs. In an upturn this model worked very well, but during a downturn this form of incentivization has fallen flat as most of the strike prices are way above the current market prices and this seems to be drawing sharp reactions from the same community of mangers who originally designed them to make managers behave as owners. The asymmetry that one sees when it comes to making a sacrifice against making a gain is similar to the asymmetry that one sees when a person is confronted with the dilemma of taking a risk against possible losses versus taking a risk against possible gains. Human beings tend to take more risk when it comes to insuring against loss versus insuring against a gain. The world is being shaped in a different way, the banks are partially being bailed out with public funds, that would need a different leadership for the banks that would be more prudent and risk sensitive; there is a dichotomy that would need to be solved that the Basel 2 would be recommending higher capital adequacy ratios and reserves for the difficult days that would necessitate higher capital needs for banks, while the current situation would demand that the banks start lending as soon as possible and as much as possible. The risk aversion would have to be met with risk proneness. This would invariably lead us to a low credit environment for a long time as the money needed to get the toxic assets out of bank balance sheets is itself a very tricky affair. The incentives in this new environment to spur the right kind of behavior must understand this. The business models must understand this. The automotive businesses would have to confront the reality that the old model of replacing cars every four years is probably gone for quite some time. The building segment would have to understand that housing starts would remain to be at the subdued levels for a very long time. The entire consumer industry that relied on personal finance and much relaxed environment of credit would have to deal with a different situation.

When the markets are down because of lack of credit or otherwise, the top line has two kinds of pressures; the first is the lack of saleable volumes and the second is the lack of ability to get the right prices. The second one stems from the excess capacity that exists in the market. But the market will slowly adjust this capacity, as no one would like to make cash losses, those that are in the wrong side of the cost curve would have to close down. The more this keeps happening capacity over-hang would get adjusted, but at a phase lag. So there is actually no deep pressure in the market to lower prices; in a downturn like this unless someone wants to trade cash against profit in a very distressful manner. Thus being on the right side of the cost curve is of paramount importance during the downturn. Quite unknowingly what played so well during an upturn in the pricing game, the incentives fail to attract any strategic decision making process during a downturn. The supply side game stumbles upon the demand side problem. The problem actually shifts to

cost leadership or innovation. In fact even innovation stumbles as the markets targeted for such products also return a lukewarm response in a downturn. At the end there is nothing left but cost, and there is no alternative to this. Most winners know to do this well. So the incentives must be so designed that this gets a distinct visibility. Actually the gains so derived from these actions could last forever. The survival in the short term provides a strong footing to build on and product differentiation or supply chain designs that deliver a better value become effective strategies for the future. In that respect incentives designed that tries to focus on the process rather than the results could actually be more effective during a downturn. Who knows they could be extended to upturns as well and with superior results? The action by government to tame large scale risk taking that could bring in entire economies to a grinding halt is the hot topic of today. By taking these actions we could surely put to rest the long held fear of leaving the markets to decide our destiny, entirely. But the question is whether these punitive actions would serve as the right kind of incentives for spurring the current demand, the dearth of which is the root of the problem that we can hardly ignore. Expecting that increased lending again could solve all economic problems in a downturn and incentivizing those actions that help to achieve this end may not be the solution, it could just be a part. Financial instruments alone could make limited impact and there lies the challenge of incentives. Procyon Mukherjee Zurich 24th March 2009 Procyon Mukherjee lives in Zurich.