Portfolio Disclosure Oct 2014

A prompt from Rowan Simpson (16 investments, $4m) and Ben Kepes (14 investments, $500k) led me to update my own investment Portfolio page.

I’ve personally invested $780,000 into 10 companies, and have equity received as a founder in several others, totalling over 15.

Overall I’m really happy with the private investments. They are relatively small amounts of money for the companies though, as I’ve only been able to invest as I earn. The mistakes have been expensive, but the amounts were small versus more successful investments. I’ve also reinvested two or more times in some of the larger investments, where it is clear that the value is there. These reinvestments have, so far, all paid off.

It’s really hard to talk publicly about how I value my investments, as each company may have different perspectives to my own. But my own take is that I’ve seen an aggregate of over 50% IRR on the investments I’ve made, and perhaps well north of that.

With Punakaiki Fund we raised and placed about $1.5 million into four great companies, and are anticipating raising more money to invest in the next few weeks.

Posted in NZ Business

Set a high bar and never compromise

Laszlo Bock, Google’s SVP People Operations, shares three secrets for how Google recruits.

He does this at the end of a presentation he gave at a Linked-In conference for recruiters.

Action starts at 2m:40s.

Laszlo doesn’t mention it, but these are exactly the same (with a bit more science these days) “secrets” that McKinsey and many other great companies use.

Maintaining the high standards is very hard – they tend to drop over time as companies get bigger. What Google do is to keep every hiring decision (but not the interviewing work)  with a small group of very high standard setters – much like Steve Jobs set the insanely high benchmark for every Apple product. Even today Larry Page reviews every hire.

They use structured interviews, a combination of behavioural (Tell us about a time…) and situational (Case studies), and they test for general cognitive ability, leadership, Googleyness and role-related knowledge. They de-prioritise the last, liking people with more diverse role experiences, as role-related knowledge can be learned while those other experiences can bring different ways of approaching problems.

The most important issue from the video is not ever compromising on standards. This is particularly important for the first few people hired at growing companies, but it can never change as companies grow. Your new co-workers will dictate how well you perform in the future, and how fun it’s going to be. So take your time, make sure there is a alignment of values and fire fast if things don’t work out. Companies like Xero, Vend, Timely and Define Instruments* are incredibly selective and difficult to get recruited into (and they are  hiring). That’s by design, and that’s why they can stay ahead.

(I’m a direct or indirect investor in all of these)

Posted in NZ Business

Tracking the performance of the 1 hour Xero model

DISCLOSURE: I hold Xero shares. 

Last year I built a very quick and dirty spreadsheet to analyse Xero, and wrote Valuing Xero – in one hour. The article was cross-posted to the NBR, where it attracted far more comments. More on those comments below.

I used Xero’s September 013 earnings release (and only that earnings release) to feed the spreadsheet and to try to keep things simple. It was not meant to be a valuation – just a chance to show how simple it is to create a model from very little information.

With the recent release of Xero’s September 2014 results, it’s time to revisit.


I’m not an official financial advisor, and you should make your own counsel and seek advice before investing. I do urge you do do your own numbers, regardless of any external advice that you get. Also understand the difference between placing short term bets against market movements by traders and long term plays based on intrinsic business fundamentals. I, for example, generally buy and hold investments, whether public or private, as that allows the investment value to reflect long term results rather than actions of other market participants.

1: How did I do?

Lousy. I made a basic formula error in the spreadsheet, discovered by the ever-present Anonymous at NBR. Well done to him or her, and brickbats for me.

That formula error transformed the model dramatically – from showing that Xero’s value was $2-8 billion to showing that it was several times larger. Here’s what I said a bit later in the comments:

I agreed with the comment that Xero has a shot at being New Zealand’s biggest company ever by market capitalisation, and correcting the formula error did give me a lot more comfort that the model was working.

The error in the spreadsheet was at the final step, which I had clearly rushed in my haste to deliver the result within my own time limit. The summary table showing the customer numbers and the annual revenue and expense results had some incorrect references in the later years. These latter two were summed to calculate the annual EBIT, which was used to calculate the Net Present Value of Xero’s EBIT. One error (customer numbers) was a typo, and was just visual, but the other error meant that the profit in the last few years and hence the terminal value were well under-reported. Here’s what the error looked like:

The cells were each summing up 12 months worth of data, and I had simply not completed the line for revenue and expenses. The most embarrassing is the cell showing the sudden drop in customer numbers (2021 was going to be a bad year), as it looks so obvious.

In a real model I would generally have a large number of charts showing the trends, and when constructing models I often create then delete charts just to make sure the numbers are making sense. Here’s the chart that I should have created:

But I did not, and so the difference was embarrassingly large.

The first bottom line (Total value of Xero) showed total values of between $2.2 and $8.8 billion, depending on the discount rate. (I tend towards 12-15% but you can pick and choose).

Before fixing the error:

After fixing the error:

The amended version shows valuations from $13 to $67 billion with the same discount rates, and $5.5 billion at 25% – placed there for  Anon (comment 16 on NBR) who wrote that “This remains a venture capital play and I would suggest a WACc closer to 25% would be justified, especially for investors with limited exposure to high risk transactions.” What anon is saying is that he or she believes that there is considerable uncertainty in the future of Xero and that investors should demand higher returns.

The error undervalued Xero by a substantial margin.

However there were a number of other corrections suggested to the model, such as salary rates, checking the number of customers in the out years, key man risk for Rod Drury, validating versus value of incumbents that Xero is attacking, tax, inflation, capex and so on. Many of these would depress the value, but none would do so in a material way versus this simple error. At least the error was in the right direction, and it goes to show that sometimes it does pay to read the comments – even in the NBR.

2: Apart from the error, how did  the model hold up?

The good news:

The model (these numbers were the same in the corrected one) projected an annualised runrate as at the end of September 2014 of NZ$131.6 million. The actual reported number was almost exactly the same at $132.3 million (both round to $132 million). That’s very good.

The number of paying customers at the end of September was projected to be 383,000, but was delivered at 371,000, 3.2% under my forecast. I’m taking that as a win as it is within any reasonable margin of error, especially as the number had increased by 75.5% versus the year before.

The operating revenue for the six months was reported as $54.3 million, while the subscription revenue was $52 million, splitting these two out for the first time. The 2013 forecast showed $58.2 million, so the result ($54.3m) was 6.7% lower. But wait – Xero also reported that their subscription result in constant currency was $56.2 million, which is just 3.4% away from the actual. So I’ll take that as a win as well, however it’s clear that the model is now out of date and I need to adjust for the break out of subscription revenue.

It would also be nice to model the effect of currency on Xero’s results, but I’m probably not going to do that just yet. Over time the foreign-derived income will increase and provide a natural hedge for offshore costs.

The changes:

The material changes have been in the regional results – shown here by number of customers. The North America result was lower than expected, casting doubts in many people’s minds about whether or not Xero is going to win in the USA.

I’m not going to change my own position, which is that Xero will grab a substantial share of the US market, and that what we are seeing is standard for SAAS subscription companies. They seem so slow in the early years, but the effect of compound growth is that over time the numbers start getting huge. While the forecast numbers may seem a long way from a result, in reality the growth is so high that the forecast is usually just a month or two early or late. The overall number of customers, for example, was 13,000 over-forecast, but that represents just over half a month’s worth of net new customers. It’s similar to the classic lily in a pond that doubles in size each day – the lily covers just half the pond on the penultimate day.

However the US/Canada and rest of world total missed by a larger margin – 2.5 months or 17.4%. Most of that growth came from the  US market, which is now broken out showing growth from 10,000 to 22,000 in the 1 year from September 2013 to September 2014.

Xero did not report annualised revenue run rates by region, which is a pity. They did report total revenue by country for the six month periods to September 2013 and 2014 though. Here’s how the forecast (middle column) performed versus the result – only the NZ result was close, but given the adjustments above (currency, split out) it was probably a tad high while the Australia/UK growth were closer that shown.

The forecast was working hard here from the limited 2013 information, and it’s clear that some adjustments are required.

3: Summary

Here’s the file2013 spreadsheet with fixed formulae and comparisons. Please handle with care.

I’m very happy with how the model performed, but it’s also clear that it can be improved with the new information from Xero and by making some changes in assumptions in reaction to the comments. I’ve started doing this but am not sure whether I will continue – there is too much work to do. The model is ridiculously simple and still a long way from a proper financial model – so tread with absolute caution, and watch out for errors.


On Xero

I remain firmly of the opinion that Xero is destined to be a giant, and we should ask of Xero (and every recurring revenue company) not “Why will the growth continue?” but “What would make the exponential growth slow down or stop?

It’s clear that in New Zealand the growth is slowing – is this because Xero is hitting the end of the addressable market, or will we see slower adopters steadily coming across? What are the month to month additions and churn rates for the North American market, and when will we run out of customers in the Australian and UK markets? Xero addressed some of these questions in a recent release, and I intend (no promises as above) to adjust the model to take these into account.

Overall it’s a timing question – I remain confident that Xero will hit, say, $1 billion in revenue, but the question is when. The unadjusted model suggests that this particular benchmark would be close to happening in FY2018, but that will certainly slip a year or two once I make some reasonable adjustments. But does it really matter when $1 billion revenue occurs? Xero is clearly a monster in our midst.

Parting shot to another anon from last’s year’s NBR article – who was of course proven correct almost instantly.

Posted in NZ Business | 1 Comment

Don’t cough on me

It used to be acceptable to go to work or travel with a cough or the flu. That’s been changing over the last 10-20 years, and people who cough and sniffle in public are increasingly treated like people who smoke in the same places. Good.

But now the stigma of being sick in public is even worse:

Let’s be clear – if we manage this well (and we will) Ebola will be tightly contained in almost all countries.  Managing well means contact-tracking, where everyone who comes into contact with an infected person is tracked, quarantined and cared for until they are deemed safe. This is the standard technique to prevent diseases taking over, and it was also used to slow the spread of HIV. Unlike HIV, which was always contagious, and even when no symptoms were evident, Ebola is only contagious for a limited time, and only when the patient is obviously sick. So Ebola is inherently easier to manage and scaremongering aside we will all move on.

How do we know? Well if Nigeria can do it, we all can.

But even under a well managed regime it’s still going to be very awkward to be visibly sick in public places in the next while.  Here’s how I see it right now. Do you agree?

  • It’s not acceptable to go to kindergarten or school when ill. The virus will spread and one family’s inconvenience becomes magnified by a classroom and their families.
  • It’s almost completely unacceptable [fixed typo] to go to work when ill. The virus will spread and one person’s brave “work at any costs” attitude can remove tens of others from work.
  • It’s becoming unacceptable to fly or be on public transport when ill. We are in very close proximity to each other and disease will spread very easily.
  • It’s not smart to not be treated if symptoms persist.

But society is not setting us up for success here.

  • Some parents find it difficult or unaffordable to take time off work, either to deal with either their own sickness or that of their children.
  • Some people are not living in environments that allow them to get and stay well, for  example very draughty homes in winter.
  • Some employers are intolerant of workers taking “excessive” amounts of time off.
  • Airlines make it expensive or impossible to move travel dates due to sickness, forcing  travellers to take flights where they should not. (Air New Zealand’s pursers even shake the hands of essentially every business class or passenger with high status at the beginning and end of long-haul flights – something I find hygienically abhorrent and difficult to opt out of. )
  • Doctors visits still cost money, and even finding a doctor is harder than it should be.

We need to continue to collectively change our behaviour. If we do so we’ll not only slow the spread of disease, and increase our happiness, but we will also increase overall productivity.

As Employers we can be smarter about sick (or child-sick) days, making sure employees understand they have an obligation to do the right thing and stay home. This is easy to apply for businesses where working from home is a genuine option, but more important to apply in customer service businesses where staff have contact with many people.

Schools seem to be a lot smarter about dealing with sickness, but let’s also investigate other options to help kids stay interacting with the class and content while they are away at home in quarantine. Preferably without affecting the parents too much.

As a Society let’s get intolerant of housing that is subpar, building on a movement that already has momentum. And let’s increasingly look askance at people who bring their illness into public.

As above all let’s take Personal Responsibility, and I see this happening a lot more. Stay at home if you are ill, and if you absolutely must be on a plane from Auckland to Wellington at 6am on the 13th of October with your coughing and spluttering, then please strongly consider wearing a surgical mask, as is common in more and more countries.

Posted in NZ Business | 1 Comment

SellShed shedding money?

This is not how you are meant to do it: Online seller SellShed starts up

The seven-person firm has invested hundreds of thousands of dollars building a website and free iPhone app and was now on the hunt for “smart money”

A very bad sign. Rasing money just to build a website and app implies they outsourced the technical stuff (and the question is what on earth is left), and every change from now will cost money. And they will need a lot of changes.

SellShed does not have a payment engine and does not charge a commission. Howell said it instead encouraged users to contact one another and trade direct, believing its service could be funded through advertising and possibly in future by paid “premium services”

Another very bad sign is that there is no revenue model, so the flow of money would continue. A payment engine is pretty basic for this sort of thing – at the very least for premium listings of the ability to buy Facebook ads.

The app had been download nearly 5000 times in its first week,

Better – but this is a vanity metric. The next metrics to track are the actual use, number of completed transactions and, the only one that matters, revenue.

It’s sad to see a lot of money wasted. Time will tell, but the early portends, at least from this article, are not good.

Posted in NZ Business | 5 Comments

A victory for great coffee

Source: Restaurant Brands NZ Ltd 2014 Annual report

Posted in NZ Business | 1 Comment

Are you innovative enough?

I’ve been reading Peter Thiel’s excellent and short book – Zero to One. It’s resonating with a lot of people at the moment, and justifiably so. It reminds me a lot of Peter Drucker’s  Innovation and Entrepreneurship, published originally in 1984 and with very similar messages.

My own take is that the art of business is easy, but the way we talk about it keeps changing, at least superficially. I do recommend reading both of these books.

Thiel differentiates between companies making incremental changes, such as those who create new products in a category or new markets for a product, and companies creating entire new markets. Drucker did as well, saying that entrepreneurs search and take advantage of changes, and “create something new, something different; they change or transmute values.

All this this is aligned with design thinking, which, as we did at Better by Design, preaches that the goal is to capture a global niche – something Thiel also pushes. While Thiel puts end-user centred design in the incremental change bucket, and we can argue either way about that, we can all agree that nobody knows they need a product in a category that does not yet exist. The Walkman used to be the classic example of this – a product created by the strong will of Sony Co-Chairman  Masaru Ibuka, while more recently Steve Jobs was famous for doing similarly at Apple.

While Thiel is not enamoured with incremental companies, that’s not to say there are not great returns possible for founders and investors in those companies. Trade Me shareholders were, no doubt, happy with their status as “the eBay (and other sites) of New Zealand”, even as they outperformed eBay on all the vital per capita metrics.

But the outsized gains for founders and investors, the billions and tens of billions, are reserved for really disruptive companies like Amazon, Microsoft, Thiel’s Paypal, Elon Musk’s Tesla, and Apple. Here in New Zealand we have Xero and Vend, each of which attracted investment from Thiel’s vehicle here – Valar. All of these companies changed or are changing the way business is done, and once they establish a foothold the relentless adoption curve takes over.

All of these companies delivered or will deliver outsized returns to investors, and they all required leaps of faith by founders and investors both. Succeeding isn’t certain, as the acquirers of Paypal (eBay) eventually found, but PayPal still delivered years of outsized growth despite not delivering on the true promise of PayPal. And sometimes investment can be in a product that just isn’t going to take – it’s fair to say the jury is firmly out on Valar’s investment in Booktrack, but nobody can accuse Booktrack of being unambitious.

Thiel tells the dirty little secret that we all know – the ultimate business aim should be to create a sustainable global monopoly, but never to call it that for fear of being regulated. Monopolies can extract economic rent (high margins and profits) while they dominate, and that’s where the huge valuations come from. Don’t expect, for example, Xero’s average revenue per customer to stay as low as it is forever, and be wary of Xero moving horizontally into related products once they do dominate.

Thiel suggests that we can pick the great companies as the ones that can  answer the series of questions below:

  1. The Engineering Question: Can you create breakthrough technology instead of incremental improvements?
  2. The Timing Question: Is now the right time to start your particular business?
  3. The Monopoly Question: Are you starting with a big share of a small market?
  4. The People Question: Do you have the right team? (real technologists wear T-Shirts not suits)
  5. The Distribution Question: Do you have a way to not just create but deliver your product?
  6. The Durability Question: Will your market position be defensible 10 and 20 years into the future?
  7. The Secret Question: Have you identified a unique opportunity that others don’t see?

Let’s put these into my preferred way of looking at companies, which I use for Better by Capital, Punakaiki Fund and consulting work:

End User: Timing, Durability, Secret. Is the target end user well defined, and is the product delivering something distinctive and delightful that changes their life? Is the product in a long-term defensible position, and is there a (generally obvious) future development path?  I’m not so sure that Thiel has covered this fully in his questions, but I wold hope it is taken a a given – the product or service has to be great for its chosen niche.

Customer: Monopoly, Distribution: How do you get money for the product or service? How do you expand sales rapidly and sustainably?  How do you capture a global niche market?

Company: Engineering, People: Do you have the founders, staff and governance to deliver? Do you have the internal knowledge and IP, and a track record of shipping?

Investor: Do you have all of the above, and does the investment get you to the next stage of sustainability? Do the numbers work?

For me there seems to be good overlap, and I generally do cover the points raised in the questions in my own work. But the exact form of the model we use to look at businesses does not really matter – just whether it works for you and the business. I’ll certainly be adding Thiels’ questions to my arsenal, but not to replace my own model. I do commend them for people starting on new businesses or developing products.

Posted in NZ Business | 1 Comment