Author: SK

Elasticity and Discounted Pricing

The common trend among startups nowadays is to give away their product for a low price (or no price), and often below what it costs them to make it. The reasoning is that this helps them build traction, and marketshare, quickly. And that once the market has taken to the product, and the product has become a significant part of the customer’s life, prices can be raised and money can be made.

The problem with this approach is the beast known as elasticity. Elasticity means that when you increase your price, quantity demanded falls. Some products are highly elastic – a small increase in price can result in a large drop in quantity. Others are less so. Yet, it is extremely rare to find a product whose elasticity is zero, that is, whose quantity demanded does not vary with price. And even if such products exist, it is extremely unlikely that a product produced by a startup will fall in that category.

A good example of elasticity hitting is the shutting down of this American company called HomeJoy. As this piece in Forbes explains, the chief reason for HomeJoy shutting down is that it couldn’t hold on to its customers when it started charging market rates:

Not only did that kind of discounting make Homejoy lose significant money, it also brought in the wrong kind of customer. Many never booked again because they weren’t willing or able to pay the full price, which ranged from $25 to $35 an hour. Homejoy changed its pricing last year to make recurring cleanings cheaper and encourage repeat business. In response, some customers simply booked at the cheaper price and cancelled future appointments.

Based on the above explanation, it seems like subsidising customers to gain traction is a bad idea, and that a business should not be willing to make losses in the initial days in order to gain market. Yet, that would be like throwing out the baby with the bathwater, for subsidising at the “right level” can help ramp up significantly without elasticity hitting later. The question is what the right level is.

A feature of many businesses, and especially marketplace kind of businesses that startups nowadays are getting into, is economies of scale. This means that as the number of units “sold” increases, the cost per unit falls drastically. In other words, such businesses work well when they have built up sufficient scale, but collapse at lower levels. For such businesses, the thinking goes, it is impossible to bootstrap, and the solution is to subsidise customers until the requisite scale can be built up, at which point in time you can start making money.

The question is regarding the “sweet spot” of subsidy that should be given to the customer in order to build up the business. If you subsidise the customer too much in the initial days, there is the risk of elasticity hitting you at steady state, and things rapidly unravelling. If you subsidise too little, you may never build the scale.

The answer is rather straightforward, and possibly intuitive – start out by charging the price to the customer at which the business will be profitable and sustainable in the steady state. This will imply losses in the initial days, since your unit costs will be significantly higher (due to lack of scale). Yet, as you ramp up and hit steady state, you don’t have the problem of raising the price which might result in elasticity hitting your business.

What if, on the other hand, the subsidy you are giving out is not enough, and you are not willing to build traction? That is answered with the “Queen of Hearts” paradigm. The paradigm says that if the only way you can make your contract is if West holds the Queen of Hearts (talking about contract bridge here), you simply assume that West holds the card and play on. If he held the card, you would win. If not, you would have never won anyway!

Similarly, the only way your business might be long-run-sustainable is if you can generate sufficient traction at your long-run-sustainable price. If you need to drop the price below this in order to gain initial traction, it means that you will have the risk of losing customers when you eventually raise the price to the long-run-sustainable-price, which means that your business is perhaps not long-run-sustainable, and it is best for you to cut your losses and move on.

 

Now think of all the heavily-discounted startups out there and tabulate who are the ones who are charging what you think is a long-run-sustainable price, and who runs the risk of getting hit by elasticity.

On Uppi2’s top rating

So it appears that my former neighbour Upendra’s new magnum opus Uppi2 is currently the top rated movie on IMDB, with a rating of 9.7/10.0. The Times of India is so surprised that it has done an entire story about it, which I’ve screenshot here: Screen Shot 2015-08-17 at 8.50.33 pm

The story also mentions that another Kannada movie RangiTaranga (which I’ve reviewed here) is in third spot, with a rating of 9.4 out of 10. This might lead you to wonder why Kannada movies have suddenly turned out to be so good. The answer, however, lies in simple logic.

The first is that both are relatively new movies and hence their ratings suffer from “small sample bias”. Of course, the sample isn’t that small – Uppi2 has received 1900 votes, which is 3 times as much as its 1999 prequel Upendra. Yet, it being a new movie, only a subset of the small set of people who have watched it so far would have reviewed it.

The second is selection bias. The people who see a movie in its first week are usually the hardcore fans, and in this case it is hardcore fans of Upendra’s movies. And hardcore fans usually find it hard to have their belief shaken (a version of what I’ve written about online opinions for Mint here), and hence they all give the movie a high rating.

As time goes by, and people who are not as hardcore fans of Upendra start watching and reviewing the movie, the ratings are likely to rationalise. Finally, ratings are easy to rig, especially when samples are small. For example, an Upendra fan club might have decided to play up the movie online by voting en masse on IMDB, and pushing up its ratings. This might explain both why the movie already has 1900 ratings in four days, and most of them are extremely positive.

The solution for this is for the rating system (IMDB in this case) to pay more weightage for “verified ratings” (by people who have rated more movies in the past, for instance), or remove highly correlated ratings. Right now, the rating algorithm seems pretty naive.

Coming back to Uppi2, from what I’ve heard from people, the movie is supposed to be really good, though perhaps not 9.7 good. I plan to watch the movie in the next few days and will write a review once I do so.

Meanwhile, read this absolutely brilliant review (in Kannada) written by this guy called “Jogi”

One Rank One Pension – some thoughts

There has been a lot of debate of late on whether veterans should be moved to a “one rank one pension” system. I won’t bother explaining the whole deal here, I’ll let you read this brilliant post by Ajay Shah about the numbers behind the move. Now that the quant has been outsourced, I can put forth my “qualitative” arguments.

I’m not a fan of this One Rank One Pension (OROP) move. I’m not against paying our soldiers, or veterans, well – I think it must definitely pay above market rates for the skills required for the job. Yet, I think OROP is a “one delta” solution to the problem (previous post here about government’s one delta thinking on agriculture), and can lead to massive unfunded liabilities.

The problem with any kind of pension scheme is that you create liabilities today that need to be funded later on. And at a later date these liabilities might become unserviceable. From this perspective, it is important to try and fund any future liabilities today, or at least have a handle on the precise magnitude of liabilities required. OROP, being “inflation indexed” (that’s Ajay Shah’s nice model to look at it), doesn’t allow for proper budgeting and long-term planning.

It is precisely due to this budgeting issue that the government moved most of its incoming employees to the New Pension Scheme (NPS) in 2004. NPS, unlike previous pension schemes, is a “defined contribution” scheme, where your pension is paid out of a corpus you create by your own saving. From an accounting perspective, it moves liabilities from tomorrow (pensions) to today (higher salary to fund the contributions), and is an excellent move. And there is no reason for it not to apply to the armed forces.

Most of the arguments being made in favour of OROP are emotional (“how can you deny our veterans money” etc.), and not well backed up by logical or economic reasoning. One of those is that lower-level military persons retire when they are 35, and hence need a “one rank one pension” (which I absolutely fail to understand). While I understand that the rigours of the role imply early retirement, I don’t see why defined contribution doesn’t solve the problem. It will have to be matched with higher salaries (to fund the contribution required for a long lifetime of retirement), but that implies liabilities are funded today, which is superior to pushing liabilities under the carpet for future  generations.

The thing with NPS is that it cannot be pushed retrospectively, and hence can apply at best to all forthcoming hires. We still need a solution for the existing employees and veterans, who are already on a defined benefit scheme. Yet, the important thing to consider is that the beneficiaries should be divided into three categories – current veterans, current servicemen and future servicemen, and we should find separate solutions for the three.

It might be argued that without defined benefit pensions, it might be hard to attract talent for a high-risk job like the military, and that is why we might need OROP. This is where the “derivative thinking” comes in. The thing about a job in the military is that there is a higher-than-civilian risk of losing life or limb. The solution to that is not blanket higher compensation – it is risk management.

What we need is generous death and disability insurance for our military, and this too should be purchased by the military from a professional Life Insurance firm. A generous insurance package can help mitigate the risks to life borne by military personnel, and should be sufficient to attract necessary talent. The purchase of such policies from professional insurers is important, for you don’t want the military to be doing an actuary’s job. More importantly, such a purchase will push liabilities to today rather than to tomorrow, and the last thing an army will want during the time of war is increased expenses on account of insurance.

The current debate about OROP has opened the door for a complete overhaul of military compensation. The government should jump at this, rather than simply get bullied by veterans’ groups. As Nitin Pai argues in this editorial in the Business Standard, compensation is an economic decision and should be made based on economic (and financial) reasoning, not based on emotion.

Finite and infinite stories

Stories in books or movies are “finite” in that there is a defined end-point. Real life, on the other hand, has to go on.

Recently I started reading a book called “Finite and Infinite Games”. I’m barely through the Kindle sample, so can’t comment much on the book, but I want to talk about a related concept – finite and infinite stories.

An important feature of the story is that it is “finite”, and has a fixed ending. For example, if you take Lord Of The Rings, the story is primarily concerned with whether Frodo can destroy the ring by taking it to wherever it came from before Sauron can get his hands on it. Once either the ring is destroyed or Sauron gets his hands on it, the story is essentially over, and doesn’t concern about any subsequent events.

Thus, as you plough through either the books or the movies, you condition yourself to the story “ending” at one of these two finalities. And in this particular story, considering that both of these are epochal events, all characters have a horizon no longer than the time required for one of these two events to happen. In other words, most books and movies are “finite stories”, and efforts in those stories are optimised for such finiteness.

Real life, however, is different, in that it is “continuous”. Whatever happens, in most cases, life simply goes on, and hence you need to optimise for the long term. Let’s say, for example, that you are going through a tough time at work and want your current assignment to end. And while you are at it, you look upon your life as a story, where the success or failure of your current assignment is an epochal event. Consequently, you will use a strategy that optimises your performance until this epochal event.

And then this event happens. Let’s say the assignment is a success. Then, life has to move on and another assignment gets thrown at you. Except that you’ve thrown all you had at the previous assignment, and now have no energy left to deal with this one.

In that sense, real life is like an “infinite story” (though death adds a degree of finiteness to this). However epochal certain events seem, unless they are life-threatening, one ought to think for the long term and plan for beyond the event. For unlike in the books or the movies, the story never ends.

Understanding Stock Market Returns

Earlier today I had a short conversation on Twitter with financial markets guru Deepak Mohoni, one of whose claims to fame is that he coined the word “Sensex”. I was asking him of the rationale behind the markets going up 2% today and he said there was none.

While I’ve always “got it” that small movements in the stock market are basically noise, and even included in my lectures that it is futile to fine a “reason” behind every market behaviour (the worst being of the sort of “markets up 0.1% on global cues”), I had always considered a 2% intra-day move as a fairly significant move, and one that was unlikely to be “noise”.

In this context, Mohoni’s comment was fairly interesting. And then I realised that maybe I shouldn’t be looking at it as a 2% move (which is already one level superior to “Nifty up 162 points”), but put it in context of historical market returns. In other words, to understand whether this is indeed a spectacular move in the market, I should set it against earlier market moves of the same order of magnitude.

This is where it stops being a science and starts becoming an art. The first thing I did was to check the likelihood of a 2% upward move in the market this calendar year (a convenient look-back period). There has only been one such move this year – when the markets went up 2.6% on the 15th of January.

Then I looked back a longer period, all the way back to 2007. Suddenly, it seems like the likelihood of a 2% upward move in this time period is almost 8%! And from that perspective this move is no longer spectacular.

So maybe we should describe stock market moves as some kind of a probability, using a percentile? Something like “today’s stock market move was a top 1%ile  event” or “today’s market move was between 55th and 60th percentile, going by this year’s data”?

The problem there, however, is that market behaviour is different at different points in time. For example, check out how the volatility of the Nifty (as defined by a 100-day trailing standard deviation) has varied in the last few years:

Niftysd

As you can see, markets nowadays are very different from markets in 2009, or even in 2013-14. A 2% move today might be spectacular, but the same move in 2013-14 may not have been! So comparing absolute returns is also not a right metric – it needs to be set in context of how markets are behaving. A good way to do that is to normalise returns by 100-day trailing volatility (defined by standard deviation) (I know we are assuming normality here).

The 100-day trailing SD as of today is 0.96%, so today’s 2% move, which initially appears spectacular is actually a “2 sigma event”. In January 2009, on the other hand, where volatility was about 3.3% , today’s move would have been a 0.6 sigma event!

Based on this, I’m coming up with a hierarchy for sophistication in dealing with market movements.

  1. Absolute movement : “Sensex up 300 points today”.
  2. Returns: “Sensex up 2% today”
  3. Percentile score of absolute return: “Sensex up 3%. It’s a 99 %ile movement”
  4. Percentile score of relative return: “Sensex up 2-sigma. Never moved 2-sigma in last 100 days”

What do you think?

Grofers, BigBasket and the Lack of Systems Thinking

Last week I wrote this post about why Grofers is not a sustainable and scalable business. The basic point was that goods they sell undergo both high inventory cost (having been stored in a retail store) and high transport cost (delivery).

The most common response to the post was that my claim was wrong because “Grofers doesn’t store any inventory but only delivers”. And it was not unintelligent people who said this – I counted at least three IIM graduates who made this claim on Twitter (ok if that statement gives the impression that I think that all IIM graduates are intelligent, so be it. I don’t disagree).

While their claim is correct, that Grofers doesn’t store any inventory but only delivers, the problem with their line of attack is that they are looking at it from a very localised perspective and not looking at the bigger picture.

A similar problem can be seen in this post on TechCrunch announcing BigBasket’s latest round of funding. Relevant section here (hat tip: Rohin Dharmakumar):

Challenges faced by BigBasket include the grocery industry’s low margins, the cost of adding new delivery staff, and the fact that it carries its own inventory. This allows BigBasket to offer a large selection, but also means it has more overhead than hyperlocal services that partner with existing merchants and needs to more time to prepare before expanding into new cities. (emphasis added)

Catherine Shu, who wrote that piece, might be right in claiming that Bigbasket carries its own inventory. But she is wrong in claiming that it is a problem, for Bigbasket is in a completely different business compared to those hyperlocal services, and in my opinion in a superior business. The carrying of inventory is a feature rather than a bug.

What the twitter comments on my post on Grofers and this piece on BigBasket illustrate is the lack of “systems thinking”. People are great at looking at localised problems, and localised “point solutions” to these local problems. What is not so intuitive is to look at a particular problem as part of a bigger picture and in a more holistic fashion.

Grofers itself may not carry inventory, but the goods it ships would have been part of inventory of some retailer. So while Grofers may not directly incur this high inventory cost, someone along the chain (the retailer in this case) does, and that means there is less money for Grofers to play around with and make a margin.

BigBasket, on the other hand, carries its own inventory and this inventory is aggregated at a much higher than retail level. This implies that the inventory costs for BigBasket are significantly lower than any retailer (since aggregation leads to lower inventory costs). And this inventory cost thus saved can help BigBasket make higher margins. It also allows them to serve the “long tail” to the customer cheaply, something Grofers may not be able to do if no shops in the customer’s vicinity stock such products.

The problem with localised thinking is that it leads to localised solutions, and local optimisation. Optimising locally at different points in a chain makes it harder to optimise at a system level.

Using all available information

In “real-life” problems, it is not necessary to use all the given data. 

My mind goes back eleven years, to the first exam in the Quantitative Methods course at IIMB. The exam contained a monster probability problem. It was so monstrous that only some two or three out of my batch of 180 could solve it. And it was monstrous because it required you to use every given piece of information (most people missed out the “X and Y are independent” statement, since this bit of information was in words, while everything else was in numbers).

In school, you get used to solving problems where you are required to use all the given information and only the given information to solve the given problem. Taken out of the school setting, however, this is not true any more. Sometimes in “real life”, you have problems where next to no data is available, for which you need to make assumptions (hopefully intelligent) and solve the problem.

And there are times  in “real life” when you are flooded with so much data that a large part of the problem solving process is in the identification of what data is actually relevant and what you can ignore. And it can often happen that different pieces of given information contradict each other and deciding upon what to use and what to ignore is critical to efficient solution, and the decision is an art form.

Yet, in the past I’ve observed that people are not happy when you don’t use all the information at your disposal. The general feeling is that ignoring information leads to a suboptimal model – one which could be bettered by including the additional information. There are several reasons, though, that one might choose to leave out information while solving a real-life problem:

  • Some pieces of available information are mutually contradictory, so taking them both into account will lead to no solution.
  • A piece of data may not add any value after taking into account the other data at hand
  • The incremental impact of a particular piece of information is so marginal that you don’t lose much by ignoring it
  • Making use of all available information can lead to increased complexity in the model, and the incremental impact of the information may not warrant this complexity
  • It might be possible to use established models if you were to use part of the information. So we lose precision for a known model. Not always recommended but done.

The important takeaway, though, is that knowing what information to use is an art, and this forms a massive difference between textbook problems and real-life problems.

Palani

I was reading Shoba Narayan’s excellent piece in MintOnSunday about the Palani temple when I was reminded of my own trip there back when I was a kid, so thought I should write about it.

The memories are extremely hazy, for I was a really small boy back then (I don’t even remember how old I was). It was a strict pilgrimage, consisting of two overnight bus journeys, and the only purpose of the trip was to visit the Palani temple.

There was some religious context to it. Apparently my parents had visited the temple some time before I was born, and had promised to return had some condition been satisfied. I don’t remember the exact condition (though the fact that I’m named Karthik has something to do with this, I know) but apparently it had been satisfied, and so off we went to fulfil the “harke”.

I remember taking a Tamil Nadu State Transport bus. I don’t think I was old enough for them to take a ticket for me, so I didn’t get my own seat. But then my father spoke to some people across the aisle and found that they were scheduled to get off at Krishnagiri, after which we crossed over to the three-seater, and I remember sleeping across my parents’ laps.

We reached Palani in the morning and checked into some random hotel. I don’t remember much of what happened there. I remember going to the temple sometime during the day. There was a cable car, if I’m not wrong, to go up. I don’t remember if we took it.

Shoba’s piece is about the Prasad at the Palani temple, but I don’t remember any of it. All I remember is going to some vibhuti (sacred ash) shop there to buy some vibhuti. And I remember the shopkeeper telling us that whatever we bought, we would only get half of it after the pooja was done. Finally my parents, after some deliberation, settling on buying one (largish) packet of vibhuti. I remember taking home half of that, and it satisfying our vibhuti needs for several years after that.

As I said right up front, this is one of my least memorable trips from my childhood. All I remember is the bus. The shady hotel. The steep flight of stairs to get to the temple (Shoba writes about that, too). The cable car. And the half packet of vibhuti. I have no clue what we ate. I think there were people there in Palani who spoke Kannada, but I’m not so sure. And I remember taking another overnight bus back (this one being empty enough that I could sleep across my parents’ laps for the full journey).

Why Grofers is not a sustainable business

When I meet acquaintances for “gencus” nowadays, one of the things we somehow end up talking about is the startup world and inflated valuations of some Indian tech-enabled startups. The favourite whipping boys in any such discussions are food delivery companies such as Swiggy or TinyOwl and grocery delivery startups such as Grofers.

All three aforementioned companies have raised insane amounts of money and are making use of these insane amounts of money to poach employees at inflated valuations. They are also launching significant “above-the-line” advertising campaigns making use of the funds they are flush with. Yet, there is one fundamental concept that indicates that these companies are not likely to go far.

The whole idea of e-commerce is that you trade inventory costs for transportation costs. In “traditional” offline retail, transportation costs are low, since everything is transported in bulk, up until the retail store. In exchange for this, there are significant inventory costs, since inventory needs to be stored in a disaggregated fashion (at each retail outlet) pushing up uncertainty, and thus costs.

E-commerce works on the premise inventory is held in an aggregated fashion thus pushing costs down significantly (especially for “long tail” goods). In exchange, the entire transportation supply chain happens in an expensive “retail” manner. Thus, you save on inventory costs but incur transportation costs.

The problem with businesses such as Grofers is that they incur both costs. First of all, since they rely on picking up goods from retail stores, the high inventory cost is incurred (the hope is that retailers will give Grofers bulk discounts, but that is capped at a fraction of the margin that retailers make). And then, since Grofers transports the item to the customer’s location, retail transportation cost is incurred (whether it is directly paid for by the customer or by Grofers is moot here, since it has the same effect on prices and volumes). Thus, Grofers incurs costs of inefficiencies of both online and offline retail, and is thus a fundamentally unsustainable business.

It can be argued that Grofers offers a degree of convenience that you pay Grofers rather than incurring the cost yourself of getting the goods from the shop. This has two problems, though – firstly, a large number of small and medium retailers in India anyway offer free home delivery (and take orders by phone). Secondly, the cost incurred by Grofers for delivery is a transaction cost and irrespective of who bears it, it results in a reduction of total volume of transactions.

In its last round, Grofers raised $35M. Given the above fundamental inefficiency in its model, it is hard to see the business being worth that much in the long term.