Thursday, 29th Aug, was a tough day for us as a business. Our vendor-managed Order Management System (OMS) had a 30+ minute downtime caused by a single anomalous order, a new scenario they were not familiar with. We investigated the issue and published a postmortem report with technical details on the same evening, which seems to have been glossed over amidst the viral coverage and the trolling that followed. We also shared the report with the exchanges and SEBI.
The media coverage was staggering. We trended on social media for the wrong reasons. There was a panel discussion on a TV channel inciting their audience to tag and tweet at us. Our competitors jumped on the opportunity, as they often do, touting their own “bug-free” and “0% downtime” services, which of course is an impossible premise.
It was a very stressful event for us as at Zerodha. What caught my attention the most though was the slew of ridiculous conspiracy theories about us, mostly from sources that have no insight into how stock broking works or the strict regulatory framework governing it. Even the above-mentioned TV panel posited a conspiracy theory, which was quite irresponsible.
I decided to write this post to address some of these conspiracy theories. Non-response will probably start making them wilder. Now that this is on the public record for everyone to read, I hope the theories can be put to rest once and for all.
“Zerodha is the only broker that faces technical glitches and downtime”
I have been a trader most of my life and absolutely understand our clients being upset when trades are affected by technical issues. But now, being on the other side as a broker and a technology company, I know that there is no way to run a business with zero downtime. Every business, exchanges to stockbrokers, banks to e-commerce sites, everybody experiences technical issues and downtimes. Tech giants like Google, Stripe, Cloudflare, Facebook, etc. have had multiple global outages in the last three months. Two prominent Indian banks have had constant downtimes in the last two weeks. Unfortunately for a stockbroker, especially one of our scale, the impact of downtime is accentuated manifold.
The last incident before this one happened on 27th, February 2019, six months ago. One just has to look at Twitter to see that several other brokers have had multiple incidents in the past few months. Of course, our benchmark isn’t our competition. We are an order of magnitude larger in terms of scale and more proactive on social media than them. So, the attention we get is much larger.
To give you an example, recently, one of the top five brokers in India was down for over 2 hours and had less than 10 tweets referring to the incident. We were down for 30 mins on Thursday and were trending on Twitter’s homepage and news channels were flashing our name.
As always, we continue to innovate and improve our products every day, and work hard to prevent technical issues. As I’ve reiterated several times, our low brokerage costs have no bearing on the size of our technology investments or the quality of our products.
Order Management System: Our OMS is leased from, and fully managed by, Refinitiv (previously Thomson Reuters, now owned by the London Stock Exchange). They are one of the largest exchange-empanelled OMS vendors in India powering dozens of brokers, big and small. Our in-house products like Kite and our tech stack sits on a layer above the OMS. All brokers have an OMS connecting them to the exchanges, and apart from a handful who have their own legacy systems, every other broker relies on an empanelled vendor like Refinitiv.
While Refinitiv continues to actively improve their systems, the significant majority of downtime incidents we’ve had over the years have stemmed from the OMS due to our sheer scale, which our competitors do not have. The underlying infra and backbone for the Indian stock broking industry is quite dated and these are symptoms. We’ve been focusing heavily on R&D on building future-proof systems for much larger scales.
When somebody claims to have “0% downtime” or to be “bug-free”, they are being disingenuous. Any sufficiently advanced piece of technology can have bugs. The recent global fiasco that stemmed from vulnerabilities in Intel’s processors is a good example.
Zerodha should “upgrade its server”: This is highly misinformed. Our technology stack is highly sophisticated. It consists of dozens of servers spanning multiple data centers and is built to scale to several times the amount of traffic it currently receives. Every day at 9:15 AM, several lakh people log in to Kite at the same instance. Kite streams 30+ lakh market ticks every second to lakhs of traders who are online and processes 30+ lakh orders a day. Console’s portfolio and P&L breakdown data is in the tens of billions of database rows. Kite serves billions of chart candles every day. All of this is packaged into well designed, snappy user interfaces for our clients. We’ve built all this tech in-house. When it comes to retail trading activity and scale, we’re probably one of the largest trading platforms in the world.
Our low brokerage costs have no bearing on the quality of technology, products, and R&D. There is no “server to upgrade”. What would people have to say to Google about the major outage that happened a couple of months ago?
“If you’re not paying, you’re the product”
This is parrotted in many different forms by people who don’t understand our business model. That conspiracy is that since we offer free services, we must be mining our client data and selling it to 3rd parties, especially stock tipsters. While most of our clients only invest in stocks and direct mutual funds from which we don’t earn any revenue, we do charge Rs 20 per trade (or 0.01% whichever lower) for F&O and intraday equity trades. 95% of all trading turnover on the exchanges is from intraday equity and F&O trades. We have the largest retail trading community of intraday equity and F&O traders in India trading with us.
Offering free equity investments helps us grow the capital market ecosystem in India and grow our client base. But like I’ve always said, the lion’s share of our revenue comes from the active traders, who execute 15+ lakh revenue-generating trades daily. Do the math and one can know why we are a Rs 500+ crore topline business.
As a technology company, we have kept our operating costs extremely low by developing technology for not just our users, but internally for our operations. We have the highest gross margins and are probably the most profitable retail-only stock broker in India. Both within the industry and internationally, we’re considered a fintech company more than brokerage business.
The largest VC and PE firms in the world have been keen on investing in us at billion-dollar valuations. A leading global bank, in a recent publication, ascribed a multi-billion dollar valuation to us. But, we’ve never raised external money because we’ve always been profitable and have been able to grow on the merits of our products. We’ve never advertised, which cannot be said of our competition who have large budgets for online, print, and TV advertising. The quality of our products and the goodwill we’ve earned from our clients is why we’re the largest broker in India today.
Coming back to the conspiracy theory, when we have such a high-margin, high-profit, high-valuation business which continues to grow, why would anyone in the right mind do anything which is detrimental, like selling our own client data to 3rd parties to make a few bucks, let alone to spammers and tipsters? We’ve been fighting tooth and nail to address the industry-wide problem of phone number leakage that happens at many levels. In fact, we’ve worked closely with SEBI in drafting the Cyber Security Framework that lays out strict guidelines for client data privacy and protection.
As a broker, like all brokers, we monitor aggregate client positions as a part of our risk management. This is mandated by regulations. We also have to evaluate every single trade from our clients against dozens of complex rules and conditions prescribed by the PMLA (Prevention of Money Laundering Act) laws to monitor illegal and fraudulent trading behavior. We do all of this in-house within our systems and perimeters. No client data or trades are sent out of our systems. Every broker is required to do this and those who do not are non-compliant with laws and regulations.
My recent post on Yes bank, where I shared aggregate, macro-level statistics that we generate as a part of our risk checks, incited a lot of conspiracy theories. We don’t profile or mine, let alone sell our clients or trade data (which would, of course, be in violation of numerous regulations). We don’t have a marketing or “data science” department that looks at financial data to target or sell anything. In fact, we don’t do any user targeting at all. One can go to the websites of exchanges and depositories to find richer data and statistics on market behaviour.
“The technical issues are deliberate and are for market manipulation”
This has got to be the wildest, silliest, and the most ludicrous conspiracy theory of all. That a heavily regulated financial business (regulated and audited by 3 stock exchanges and SEBI) will deliberately bring down its systems, cause their customers anguish, damage hard-earned goodwill and reputation, lose revenue, violate laws and regulations, all to “manipulate markets” by “placing opposite trades” is so wild and absurd, I don’t even know what to say.
“Brokers can make money when clients lose money, i.e., by broker counter trading”
Here is something startling—on an overall basis, active F&O and intraday equity traders lose more money to trading costs (STT, Stamp duty, and other charges) and impact cost (money lost due to the bid-ask spread), than actually to the markets. This is likely to be the same across all stock brokers.
If a broker does keep taking opposite trades on the exchange for their client trades, the broker would end up losing significant money eventually on the impact cost alone. Moreover, this would be noticed by the exchanges and regulators, who not only do audits but have realtime monitoring systems. Also, common sense would dictate that no smart broker would force their clients into losses and out of the markets, especially one that does not advertise and relies on clients’ word-of-mouth referrals for growth.
Also, in India, every single order placed by clients hits the exchange systems in real time, where they are matched. This is different from many other markets where orders may be matched and filled by non-exchange intermediaries (or “darkpools”).
The only place where this sort of activity is possible are places like illegal CFD (Contract for Difference) and binary options trading platforms that essentially take counter trades to every trade taken by clients, as there are no exchanges involved. There, when a client loses, the platform earns. Since there are no trade or impact costs, such platforms make profits if all the clients on an overall basis lose money. These platforms are banned in most parts of the world, so they are usually based out of tax havens like Belize, Malta, etc., and entice people to trade by offering extreme amounts of leverage using online ads. Every once in a while, when there is a black swan event and a group of traders earn significant amounts of money quickly and much more than those who are losing, these CFD / binary trading platforms go bankrupt. The last really big one was in 2015 when Swiss franc was unpegged and it soared 30% in one day.
Prop desk: Like almost every other broker, we do have a proprietary trading desk. We can’t be building a brokerage business with so much passion and conviction if we didn’t believe that profits can be generated from the markets. Most of what we trade are low-risk positional delta-neutral strategies and investments in stocks and bonds. What we do as a prop desk has no effect on our clients’ trades.
Also, brokers are required to upload detailed reports on self and client fund splits at the end of every trading day. Not once in the last 9 years of our business has our prop positions exceeded margins beyond our own funds. We are audited frequently by exchanges and SEBI and have never had any violation with regards to ours or our clients’ funds and prop trading.
Update: Starting September 2019, we have stopped all active trading strategies on our prop desk at Zerodha to allay any concerns of either utilising client margins or of prop positions going wrong. Just treasury operations mostly in fixed income instruments now.
This is one of those seemingly technical terms that is thrown around a lot but has no substance. This particular conspiracy theory was discussed on the TV panel mentioned earlier.
I have been in the business of trading and dealing with retail traders for two decades now. I’ve personally interacted with tens of thousands of traders in person, over e-mail, on online communities, and on our very active forums like TradingQ&A, Z-Connect, and Varsity. I can say with great certainty that the single biggest reason why retail traders lose money is by trading actively with no stop-loss to the trading strategy. Many claim to make a mental note of the stop-loss, but only a few actually place stop-loss orders. And the majority of those few who do, cancel or modify as the price starts closing in on the stop-loss. So, in reality, the significant majority never really have a real stop in place when trading. So if there is no stop, what can someone really “hunt”?
Secondly, over 80% of all trades today are on extremely liquid Nifty and Banknifty contracts. Assume Nifty is at 11010, and you get to know that at 10990 there are a lot of stop-loss orders. What can you do about it? Somehow magically take the market down by 20 points, fill those orders, and then take it back up? Even if someone did have deep enough pockets to move the price to a stop, how could this possibly be a profitable trading strategy in a market where there are lakhs of traders with different views, big and small, where no one person can control the direction of the market?
The only place where “stop-loss hunting” or forcing clients out of positions is profitable is again on a CFD / Binary trading platform, where the platform is the counterparty to every client position. The platform can potentially hit any manual stop in place or push the price to the extent where the risk management system auto squares off positions for insufficient margins. This is the reason why most of these platforms give extreme 100 to 200 times leverage to trade.
On a regulated exchange platform, this isn’t possible. Exchanges also have alert mechanisms that trigger anytime the buyer and seller on both entry and exit trades are the same clients. This immediately gets escalated and the broker is required to provide an explanation. There are severe penalties and regulatory ramifications for such activities.
Tick by Tick (TBT) feeds: Finally, stock exchanges provide what is called tick by tick (TBT) data feed where you can track every single order, trade, any modification happening on the exchange. This data contains details of trades and pending orders at the exchange level, across all brokers. So if someone did have a strategy like “stop-loss hunting”, trying to figure out retail orders, etc, the best option would be to subscribe to exchange TBT data feed. It’s quite complex and expensive, but available to anyone, and is offered by almost all Indian and global exchanges.
“Zerodha benefits by restricting trading on OTM options”
Our low brokerage and superior platforms have caused a significant increase in our F&O business. But since F&O business can bring significant systemic risk as well, exchanges have a restriction on how big a broker can be. No single stockbroker is allowed to hold more than 15% of open interest (OI) on any F&O contract. Once this limit is hit, exchanges automatically shut down trading, blocking all fresh entry trades and allowing exit trades for the broker and all their clients until the OI reduces. This happens automatically, and in realtime, and obviously, is a plight no broker would want.
With the introduction of weekly options on Banknifty last year, retail activity increased significantly and our OI suddenly moved towards the maximum cap, especially intraday. What makes it worse, as I explained earlier, is that most people don’t keep stop-losses and end up holding lots of worthless long OTM options when markets move. The only practical solution is to restrict the contracts in which a client can trade. Hence we stopped allowing buying deep OTM options (we allow shorting) which can add OI significantly with small amounts of money.
This is the real reason and we have explained this many times before. No conspiracy here.
On average, we have over 70,000 daily rejected Banknifty option orders because of this restriction. If you were to multiply this by the Rs 20 / trade we charge as brokerage, you’d guess the revenue loss the business suffers daily because of this.
Our prop desk doesn’t take any Banknifty positions to make the OI limit worse. However, we have realized that this restriction hasn’t necessarily been bad for traders. It has saved our clients a lot of money as buying OTM options is generally the riskiest trades to take; almost like buying lottery tickets.
“Mock investors; encourage active trading”
My recent post on what not to do when trading (with Yes Bank as an example) went viral. While most appreciated the message, the downtime incident suddenly had quite a few people trolling me saying that I was trying to mock equity investors because they don’t pay brokerage and that I wouldn’t write something similar about traders who give us brokerage revenue.
To clarify, I think almost all of us are traders. Buying a stock with an intention to sell in any time frame, or deciding to hold it until price appreciates to the buy price when the price goes against, is not really an “investment”, but a trade. Making an investment is like buying a house with the intent to live—you do it for reasons other than just price appreciation. Now if it is a trade (intraday, short term F&O, or even a longer-term equity one), there are a bunch of rules that need to be followed to protect oneself, and to improve the odds of winning, which is what I wrote about in the Yes Bank post. The trends of overall retail % holding in a stock and approximate average price shared can be figured out by looking at the historical shareholding pattern of every company publicly available. Nothing confidential there; just aggregate statistics.
Misselling: We do not have a single person on our 1000+ team that has a brokerage revenue or commission target. No advisors or relationship managers. Not sure how many other brokers can claim that. When there are revenue targets to be met, misselling happens.
When customers ask, we generally point them towards direct mutual funds (from which we make no commissions or revenue). We do not push any high-risk products. We never send out clickbaity F&O or stock tips inducing people to trade. We have spent thousands of hours building content and answering queries which are mainly focusing on helping people do equity and MF investments. TradingQ&A.com is one of the most active capital market forums in India, and Varsity is the biggest such open educational resource in the world, just behind Investopedia.com. The spammy marketing and stock tips, that’s just not in our DNA. I doubt if the entire stockbroking industry in India combined has as much content and engagement as we have on our free and open educational initiatives.
We’ve never taken the active advisory route because we’ve spent all this while creating the best self-service trading platforms out there. Our advisory has all been in the form of education. In the near future, we’ll start offering tools to our clients that will help them take safer trading decisions.
We’ve grown to the No 1 position with zero marketing and gimmicks, purely through the merit of products and services and the support of our clients. We’ve
– disrupted the Indian stock broking industry and set a high benchmark for technology and platforms
– brought investing costs down to zero and have made trading costs negligible.
– democratized financial and stock market education by creating original content and giving it away for free and running large public community platforms.
– invested heavily in improving the capital market ecosystem in India—10 Fintech startups funded so far.
– achieved 1 million active monthly retail clients who trade and invest (source: UCC stats released by NSE), the highest for any broker in India, ever.
Alexa says that we’re the 76th most popular website in India (and the 1031st most popular website in the world). All this without ever spending a single Rupee on advertising or having a marketing team.
We have, thanks to the support of our clients, built a very valuable and successful business. We have zero debt or any external borrowings. Almost everyone on our team holds a stake in the business and most of our networth is tied to the business remaining valuable. There is no truth to the conspiracy theories being spread. The level was surprisingly intense this time around, and hence, this long post.
As always, we’re working on some major product announcements. Stay tuned.