May 29, 2015

Game Theory, Nash Equilibrium & Grexit

John Nash died last week. Someone asked me about the significance of his work and the so-called Nash Equilibrium & Game Theory which won him the Nobel Prize. It made me think a while.  Lot of economic theories are just theories with very little practical applications. Is this game theory like that ? 
Much of the geo-political situations , competitive decisions in many industries can be modeled by applying game theory principles. If you model the problem correctly, this can give you pretty good insights into your competitor's strategy & give a lot of structure & logic to your decision making . I could think of no better example than the looming possibility of Greek default & subsequent exit that media has labelled as the Grexit question
But before that, how do you translate / model a problem into a Game Theory situation where we can apply the Nash Equilibrium principles . I stumbled upon a structured approach illustrated very well here 
Its important to understand the modelling part here . You need to model the Co-operate / Not Cooperate aspect into strategies / positions here. You maybe tempted to model Greece's strategy as 'Default on Debt' / 'No Default'. But that,s actually the result of a competitive position , not the position in itself. And once you have modeled the competitive strategies, you need to assign pay-offs ( which can be just a relative number ) to those .
These payoff values are debatable. But as long as those are relatively OK, that serves the purpose ( unless you are getting into the next level of probabilistic calculations there ) 
Now, Identify Greece's best strategy to EU's ' Restructure the debt'  position. That is always to do the fiscal reforms. If you don't do the fiscal reforms, you tend to risk EU's co-operation, potentially fail to repay further installments etc 
What's Greece's best response to EU's 'No restructuring stance'  ? It is still to do the fiscal reforms as unless they do that, they risk defaults, potentially fall out of Euro.
Whats EU's best response to Greece's 'Do the fiscal reform' position ? Their best response is to do the restructuring as, if they don't do that, Greece may still default on its debt inspite of fiscal reforms and  EU / IMF may miss further repayments
EU's best response to Greece's 'No reform' position ? A grexit is a position that everyone wants to avoid at all costs. So, EU's best response is still to restructure the debt so that Greece can continue to pay the reduced debt interest & continue to be in Euro  
So, where do the strategies overlap in ?  Do the reform, Do the re-structure column.  Greece has played its part very cleverly till now. They have made the repayments so far, they have tried some reforms. They are trying to nudge EU & IMF into the Co-operate proposition now. We'll get to know by June - 5 & may be later in the month, whether they reach a ' Co-operate Co-operate' position 
Now, what does John Nash say ? " Nash equilibrium is a solution concept of a non-cooperative game involving two or more players, in which each player is assumed to know the equilibrium strategies of the other players, and no player has anything to gain by changing only their own strategy"
So, if they reach a Do the reform - Do the restructure position, no side has got anything to gain by moving off into a different position subsequently.
Let's see how it turns out ! 

May 23, 2015

Economics Of Retail Loyalty Programs

 I'm a great fan of Carrefour Loyalty programs scheme here in UAE. I think Indian retailers ought to learn a thing or two from the way they have designed it, purely from a perspective of maximizing long term profit .. I haven't seen anything close-by to the way this operates - may be its there in the developed markets of West.

Carrefour operates this loyalty points scheme - for every purchase, you accumulate varying number of points depending on the items that you buy. Once you reach 500 points, you get a 50 AED voucher.  There are two kinds of underlying point boosting schemes -  on certain days there are things like extra points / double points thats available to anyone who buys certain products  . These are presumably linked to Inventory clearing items.  So, sometimes, you get extra points for buying washing liquid , sometimes for biscuits etc.  Why I feel this is linked to inventory clearing  -  These items are usually set at the start of the week and often run out by the end of the week ( i.e, thursday ).

 The second set of 'loyalty boosters' is what it makes somewhat unique.. Based on your purchasing pattern, you get a set of coupons - like if you spend 50 dhm for meat / fish, you get 75 points etc ( within so and so date )  .  What Carrefour is doing here is to 'upsell' you or introduce you to new categories, which you usually don't spend. So, I've never bought beef from there. But I regularly get these coupons linked to those.  And these would be typically high-margin stuff for them. One of the mistakes that businesses do in designing this step is that they mix the inventory-clearing objective & upselling objective. This is a mistake that usually erodes their margins.

There is a flaw in their algorithm / logic here though. If I buy baby diapers & baby food together in a single purchase, I'm quite less likely to get a coupon which is going to incentivize me for buying either of these .On the other hand, if I buy only baby diaper in a purchase, I'm much more likely to get a coupon incentivizing me to buy baby food. Because the system apparently thinks that it is a related category - so let me prompt this customer to buy from that.  Similarly, if you buy fruits only in a single purchase, you are quite likely to get a coupon incentivizing your purchase either in vegetables / breakfast items / dairy products . I'd like to know if this hypothesis holds well in other carrefour-type loyalty schemes

So if you live near a Carrefour supermarket, I think staggering your purchase by categorizing your purchase items into largely non-overlapping categories can benefit you ( the consumer ) most

Economically, this is something that I'd call as a second-generation loyalty program. Plain loyalty programs like 'earn & burn' schemes still dominate the retail landscape in this part of the world ( Shukran points is the best example ) .  Those lack an ability to 'upsell'  / bring customer into higher margin products.  Carrefour does that nicely. This is important in retail business as unlike in services industries ( airline / hotel / dining ) etc, the marginal cost of goods is still high. So, you need to find ways to move beyond an 'earn & burn' reward scheme by catching in the customer & leveraging their spending power to the maximum.

Apr 11, 2015

Why Internet Should Not be Neutral ?

The chorus of 'open' , 'neutral' and 'free'  tend to capture the imagination of social media quite quickly. Back during my school and college days, there was this enormous campaign for 'free software' , without realizing what was 'free' in that context. Likewise , over the last month, the Net-Neutrality campaign has caught up the attention everyone . Much of this is misunderstood  & misinterpreted by the media.

The basic underlying question in this whole debate is - 'Can an internet service provider prioritize / de-prioritize certain content over the rest' ?  It is not easy to answer that unless you understand the underlying realities & implications.

There is this concept called price discrimination in Economics. It is all about pricing different consumers differently for the same service. In airline world, there are a number of different fare classes for the same sector. So, it is quite likely that the passenger sitting next to you has paid a totally different fare.  It is the dream of every business owner to do this in a perfect way - or achieve perfect price discrimination .  There is a different 'marginal utility' for any service from a customer's perspective. So, the price that he / she is willing to pay for a service could vary a lot.  So, if a producer can price the product correctly to each and every customer, he is maximizing his benefit or 'producer surplus'

Producer surplus is not just about maximizing producer's profit. It is an important component of a free-market mechanism. Adam Smith's 'invisible hand or self-directing market & unintended societal benefits cannot be achieved unless producers have a chance to maximize their surplus. Putting it in another way, unless producers are rewarded for their produce to the right level, they don't have any incentive to do more.

In a telecom/internet world, this is functioning more like a two-sided platform. There are two important players - end consumers & content/app/web providers.  ISP's like Airtel , Vodafone etc being the platform between them .

The argument put forward by proponents of net-neutrality is that if I have paid an amount for say 1GB of data, I get to decide which app / site I'm using - ISP's don't have any business to prioritize any of those.

Lets view the situation in a different way. Assume that there are only 3 applications - Facebook, Youtube, Ted  - that consume 100% of bandwidth . And lets assume that in a city, the overall customer base's bandwidth usage results in a bandwidth consumption of Youtube - 90%, Facebook - 9% and Ted - 1 %

Now, lets pick three consumers & their ideal or desired bandwidth usage profiles being
C1  ( FB - 90% , YT - 9 %  Ted - 1% )
C2  ( FB -10% , YT - 60% , Ted 30% )
C3  ( Ted - 90% , FB - 8 % , YT - 2 % )

For C3, the net-neutral world is a nightmare.  Owing to the   overall bandwidth usage distribution, his favorite content never gets delivered properly.  So, by paying an identical fare for the data plan, isn't he cross-subsidizing the population who are just hogging up the bandwidth by consuming FB & YTube?  Neither are C1 & C2 getting the ideal data consumption plans.

Can we imagine a system where C1, C2 & C3 are able to draw up a data plan based on their own bandwidth usage patterns ?  Essentially, we are talking about the so called 'anti-neutral world' here

From an ISP's perspective, this is all about creating certain priority lines and making each consumer & app provider or maker pay for the bandwidth usage as per their desired needs.  In a 'Net-Neutral' world, neither consumer , nor producer surplus is achieved - thereby making it an inefficient market infrastructure.

There is an underlying catch / danger here -  Can a company - say facebook -owing to their dominant market power buy the entire bandwidth capacity with Airtel and restrict the entire population to facebook only ?  I'd say, in  a monopolistic world, it would have been possible. But not in an ultra-competitive industry like telecom.

Would this lead to a world where, if you are a new App-maker, you would find it difficult to reach the consumers -  Maybe for a high bandwidth consuming application. But I'd say again that if your App is good, there would always be a segment of demand & that demand will find its way through the market

It is far better to live in an anti-neutral internet than in the present scheme of things where 90% of the available bandwidth is consumed by pirated movie downloads & streaming services without them paying their proportional bit for the bandwidth usage

For another interesting take, read Tim Harford's article on this