Michael Halls-Moore

My Preferred Business Model

Sunday, 11th May 2014 - 0 Comments

In this article I want to discuss the rationale for my favoured business model, which is based around high-margin, high-probability, e-commerce single item or subscription payments in education and information.

Let's describe exactly what the business model is and why it works well. Selling information products online has an extremely high-margin. This means that a significant portion of the direct revenue from sales translates into profit. In my e-commerce businesses I am typically running with a 95% margin (gross of taxation). This means that after accounting for all expenses (excluding taxation) needed to sell items in an ongoing fashion, I recoup 95% of the revenue. This is extremely high when compared to traditional "bricks and mortor" businesses.

Selling educational products is also what I consider a high-probability business model. What do I mean by this? Perhaps the best way to describe such a business is to consider an opposing model. In the technology sector it is common to fund a business with equity often via angel finance or venture finance. The basic approach is to create a situation of extremely rapid growth in order to attract later stages of venture finance with rising business valuations in the hope of a liquidity event such as an Initial Public Offering (IPO), which is a listing of private shares on publically traded markets or a corporate acquisition by a larger firm, which provides a mixture of liquidity and/or employment to founders and a return to investors. Such a business model comes under the banner of high-risk, high-reward capital gains investing. This generally has a low-probability of return, as shown by lots of data on these sorts of businesses.

Conversely, selling information products is high-probability in that a relatively predictable stream of revenue can be estimated by virtue of determined market size, conversion rates and "average basket sizes" of products. Such values do not change too much and a relatively homogeneous among e-commerce verticals. This ensures that with sufficient traffic generation, a relatively consistent revenue stream can be generated. In the capital gains scenario described above, a lot more comes down to picking a market that ensures high growth and being able to leverage that growth via investor relationship management.

Despite the low probability aspect of high-risk capital gains investing, it still attracts a large amount of people, possibly due to a mixture of observation bias ("everybody can make a million") and enjoyment of the lifestyle that comes along with such business development. I personally prefer the high-probability model as there is little risk (in the sense of uncertainty) involved, albeit at a far lower return.

The latter part of the business model involves e-commerce and single item or subscription payments. I personally find e-commerce an extremely worthwhile area to be involved with because you are able to sell products to a global audience. Unlike a "brick and mortar" (B&M) business, where you are fully reliant on a mixture of footfall and local advertising to attract customers, an e-commerce business can potentially reach anybody in the world who possesses an internet connection. Not to mention the fact that a B&M business will likely have a huge set of overheads, such as business rates, lease fees, utility bills, staff costs etc. This comes back to the point about high margins.

The final component of my business model is selling educational products. Why do I think education is a good market? Consider that people will spend significant sums of money on becoming educated or at least to be PERCEIVED as well educated. This can be seen when looking at university tuition fees in the US and UK. Certain courses, such as those related to finance, can charge in excess of 60,000 USD per year. Thus people are clearly willing to pay for a (good) education. The other reason that education products make sense is that they are very difficult to commoditise and hence not often directly interchangeable. Consider the following institutions: Cambridge, Oxford, Imperial College, Harvard, MIT etc. These universities have a brand value that is highly difficult to dislodge. The same is true of quality educational information. Thus one can produce a moat around the products that often ensure their continued relevance.

Such businesses are referred to (often pejoratively) as lifestyle businesses, which signals the fact that they have been set up solely to provide an income to fund their founders' respective lifestyles. This is contrast to the capital gains model described above where the business tends to become the lifestyle, either by accident or by design. The ecommerce businesses I run fall directly into this former category. I very much enjoy teaching and helping people learn, but I also enjoy working on deep technical projects with other colleagues that I respect. Running a set of businesses geared towards education allows me not only to be in direct control of my personal finances, but also to collaborate with others in areas of my choosing, rather than out of necessity for income.

Hence if you wish to gain financial independence with a minimal amount of long-term stress you should consider the high-margin, high-probability, e-commerce approach.


Learning Numerical Relativity

Wednesday, 16th April 2014 - 0 Comments

On and off over the last few years I've toyed with the idea of studying Cosmology to a relatively high level. Recently I picked up a copy of The Perfect Theory by Ferreira, which is a non-technical and highly engaging discussion on the history of General Relativity (GR), the main theoretical tool used in Cosmology.

General Relativity

General Relativity was developed by Albert Einstein after he completed his work on Special Relativity. He produced the Einstein Field Equations (EFE), which were eventually solved in certain cases and otherwise utilised by such towering figures as Arthur Eddington, Georges Lemaître, Alexander Friedmann, Karl Schwarzschild, John Archibald Wheeler, Kip Thorne, Jim Peebles, Stephen Hawking and Roger Penrose.

General Relativity is our current "best" theoretical tool for making physical predictions on a large scale as it describes how mass and energy produce curvature in spacetime, which ultimately leads to gravitation.

The complexity in GR arises due to the fact the EFE are a set of ten non-linear coupled partial differential equations (PDE). Early attempts at analytical solutions used simplifying symmetry assumptions in order to reduce the complexity of the equations. Later uses of theory predicted spacetime singularities, examples of which include black holes and even Einstein-Rosen Bridges, also known as wormholes.

Numerical Relativity

Although I have had undergraduate masters level training in mathematics, particularly in partial differential equations (PDE), and doctoral training in computational fluid dynamics, the mechanics of Relativity are significantly more subtle and complex than the Newtonian universe that I am more familiar with.

For some time I had been looking for a way to combine my hobbyist interest in theoretical physics with my prior research on high-resolution CFD methods. After reading The Perfect Theory I (re-)discovered an area of computational physics known as Numerical Relativity (NR). The idea behind NR is to take the EFE and simulate them numerically on a (super-)computer in regimes that are unsuitable for analytical solutions.

In standard (Newtonian) Eulerian fluid dynamics simulations a "grid" or "mesh" of points is used to store representative values of fluid quantities of interest, which are continually updated in time as the solution is calculated. In most aerodynamic CFD simulations the mesh geometry itself is not (often) dependent upon the underlying fluid being simulated. Also, the time step used in the simulation, while potentially dynamic (i.e. not of a fixed size) is equal across the entire grid.

In NR the situation is different as the geometry of spacetime (and thus of any potential simulation grid) is dependent upon the matter and energy distribution within the grid. Thus not only is the matter distribution constantly being simulated but the geometry of the mesh is changing at the same time. The only way to solve the equations on a computer by constructing an initial value problem is to foliate spacetime into what is known as a 3+1 formalism. Even with this approach the solution of such equations are formidable and generally require supercomputers to solve.

The reward for all this hard work is the ability to simulate some of the most extreme events, in the sense of matter-energy release, that we are aware of in the univerise. In particular, NR allows the simulation of black-hole/neutron star orbit and coalescence, which have been proposed as source candidates for Gamma-Ray Bursts (GRB) as well as simulation of gravitational radiation, also known as gravitational waves.

My Plan

I received a copy of Numerical Relativity by Baumgarte and Shapiro from Amazon today. The book is a relatively recent (2010) late undergraduate/entry level graduate text on the main background required to begin research into NR. I purchased the book because I wanted to see how relevant my prior formal training in high-resolution shock-capturing methods (HRSCM) was to the area.

Fortunately enough there was some intriguing discussion within the book on that very topic and how it applies to NR. This made me realise that it might be possible to combine my interest in theoretical physics with some prior high-end CFD research, at my own pace.

My first instinct going forward is to form a comprehensive literature review of both the prerequisite background material as well as the current research frontier. To that end I've already determined that I will need to be strongly versed in the following topics before embarking on any form of research:

Thankfully, many of the current papers in the field are released on the General Relativity & Quantum Cosmology section of the Arxiv pre-print server. Once the background reading has been finished, I'll be exploring the site for review papers on NR.

In addition I am considering setting up a dedicated blog/article site on the topic of theoretical physics in general. Given that QuantStart and my other projects are taking up a significant amount of time these days, I'll have to fit it in where I can!


What Can Real-Time Strategy Games Teach Us About Investing?

Sunday, 20th January 2013 - 0 Comments

A good many hours of my youth were spent playing real-time strategy (RTS) video games on my PC. Later on in life I often considered whether this was "wasted" time. I came to the conclusion that life was too short to be concerned about vaguely defined opportunity costs. In addition I came to realise how much these games are able to teach about resource management, cash flow and the nature of investing. Hence I was gaining an education in financial prudence, albeit while destroying an opponent on a virtual battlefield.

A critical aspect of nearly every RTS is resource management. There is often a finite supply of some highly-prized resource scattered throughout the terrain, which needs to be collected and refined to generate cash. This cash is subsequently utilised to train infantry, construct new buildings or deploy armoured vehicles, which can used to obliterate the enemy.

Early variants of these games incorporated a finite-resource-harvesting model. Famous examples are early games within the Command & Conquer franchise. This scenario involved a limited quantity of resource being scattered throughout the map. Harvesting units would patrol the areas and reduce the available quantity after transport and refinement. This encouraged aggressive expansion, as control of resources was generally the key to victory against one's opponent. Later variants allowed the resource to re-grow, albeit it a significantly reduced rate compared to the harvesting (which itself could be increased by deploying multiple harvester vehicles).

More sophisticated RTS games, an example of which is Total Annihilation, introduced the infinite-supply-rated-extraction model. Unit construction and subsequent power was not measured in absolute quantities of metal or energy, but rather in terms of resource consumed per unit time. The resource supply sites scattered throughout the terrain also provided a rate of extraction per unit time. Hence the goal was to make sure that the rates of extraction were always exceeding the rate of consumption.

More recent games introduce a hybrid of the two models where there is an initial fixed supply of a prized resource within the battlefield, but certain buildings or units can be constructed, which produce a recurring cash flow. The units and buildings varied in their tactical capabilities, as well as their yield. However, fundamentally they provided the commander of a team with a set of assets.

Often the commander of such an army is presented with the choice of how to deploy the capital received from harvesting finite resources. The choice comes down to short-term production of new units or investment in cash generating assets, which provide a hedge against dwindling resources on the battlefield. More often than not the average commander will choose the short-term benefit of additional units over the, admittedly expensive, cash-generating assets. Usually, this is the wrong strategy.

The more sophisticated approach is to invest a large fraction of the initial cash flow into construction of the cash-generating assets. In C&C: Generals, one of the aforementioned Command & Conquer games, this involves the construction of supply pads, hackers or black markets depending upon the choice of army. The real sophistication of this strategy becomes evident when the initial cash flow from these assets is further deployed into producing more cash generating assets.

The mathematically inclined among you will recognise this as a geometric, non-linear process, dependent upon time. These processes are characterised by slow initial growth but rapid expansion after extended periods of time. Thus, there soon comes a point where the cash-generating assets (such as the supply pads, hackers or black markets) are providing significant cash flows to fund huge expansion in unit construction, leading to easy victories.

It is straightforward to see how this scenario is analogous to the "real world" practice of investing. Consumers are often faced with the decision to deploy their initial resources (i.e. their cash!) towards a short term gratification, such as a daily coffee or expensive watch. In essence they are foregoing the ability to purchase cash-generating assets, such as a common stock which provides regular dividend payments.

Had these consumers instead put their money towards continual purchase of common stocks and re-invested the dividends into more common stock, then over a period of time cash would be generated automatically. This could ultimately be used to fund luxury items of the type previously described - without materially affecting the rate of new cash being provided.

One can also compare the finite nature of RTS battlefield supplies to a finite resource such as the amount of money earned as a salary. With finite resources on the battlefield, only a finite number of units could be ultimately produced. Analogously, with a lifetime salary earning, only a fixed number of luxury items could be purchased. By choosing to invest a fraction of this salary into cash-generating assets, with the aforementioned non-linear compounding, it would be possibly to purchase a far larger quantity of items - i.e. leverage a significantly larger amount of utility.

So, not only was I enjoying the thrill of military tactics, resource management and virtual destruction of my peers, I was setting myself up for a sound, long-term, investor mindset.


Investing £10,000 Over a 2-5 Year Time Horizon

Sunday, 20th January 2013 - 1 Comment

I received a great question from Robin yesterday asking about the different methods of hypothetically investing £10,000 with access to the funds within 2-5 years. The question was posed on the post about the differences between saving and investing.

The most important constraint that Robin mentioned is that the money needs to be accessed within 2-5 years. I'm going to assume that this money is needed for an upcoming life event, such as a wedding, down payment/deposit or similar. If this is the case, then I would strongly advise not putting the money into an investment class such as property, the stock market or foreign exchange/commodities speculation (i.e. via spread betting). Although Robin was interested in a cost-benefit analysis of different areas for investment, I can straightforwardly recommend only one approach: Saving.

Over the time horizon mentioned (2-5 years) all of these markets will likely suffer a reasonable degree of volatility. There might be large price swings in the market value of the asset purchased. This means that if the funds needs to be accessed rapidly then a substantial loss in the asset value might occur on sale if the investment is withdrawn at an unfortunate time. For instance, £10,000 invested in the UK stock market (perhaps via a FTSE All-Share Index Tracker fund) in March 2007 would have been worth approximately £5,500 in March 2009! That is almost a 50% drop in value in only two years. The key lesson here is that the stock market is highly volatile!

I am assuming that capital preservation is more important than any potential gains that may be made by investing in these volatile markets. Therefore my recommendation is to place the money in a high-interest bank savings account or a tax-free shelter like the UK Individual Savings Account (ISA), of which there are equivalents around the world.

The return on such an investment will be low, particularly at today's interest rates. In the UK the savings interest rate is roughly matching inflation (as measured by the Consumer Price Index), so in reality this would only preserve purchasing power, rather than increase it. Despite the "boring" nature of investing this way, the piece of mind that comes with knowing you can withdraw the entirety of your principal without undue risk (assuming no bank collapses in that time frame!) is very reassuring and will encourage a much sounder financial footing over the long-run.

If however, a liquid capital reserve has already been generated, the £10,000 is not required for any future event and a much broader time horizon is targeted (say 25-30 years), then investing (wisely) in the stock market is a very reasonable strategy, although this is a discussion for another blog post!

I hope that helps, Robin!

1 comment

The Difference Between Saving and Investing

Saturday, 19th January 2013 - 2 Comments

A fundamental idea in finance is the contrast between saving money and investing money. Even though these terms are sometimes used synonymously, they are actually very different and should never be confused. In this article I will outline the differences between the two, allowing you to get off to a great start with your investing career.

One of the key differences is due to a concept known as liquidity. This refers to the ease at which an asset can be bought or sold. Cash stored in the bank is an example of a highly liquid asset. It can be redeemed rapidly without penalty. A rental property is an example of an illiquid asset. To sell the property requires time and strong demand, otherwise a penalty is incurred in the form of reduced price for the sale.

The Differences in Saving and Investing

Saving and investing are generally defined by their differences in liquidity, as well as the expected return earned on any potential asset.

Saving involves placing cash into a bank current account, bank savings account or into a Cash Individual Savings Account (ISA). At the current base rate of interest in the UK, these accounts are unlikely to generate an attractive return on investment. The goal of saving is to ensure that you have access to cash for the day-to-day situations. In particular - emergencies, an unexpected redundancy, a house deposit, university tuition fees.

Since cash is highly liquid, it can be deployed easily in these situations. The disadvantage of saving cash is that the interest rates offered by banks/ISAs is rather low. These rates often only match the current inflation rate. In another post I have advocated creating a cash liquidity buffer.

Investing differs substantially from saving. It involves deploying cash to purchase an asset that generates a return over a particular time-horizon. This return can take the form of a capital gain, which is when the sale price exceeds the purchase price. It can also take the form of a dividend or coupon, which is a quasi-consistent variable payment, usually the cost of lending money or a share of the profits from a business.

Investments are often (much) more illiquid than cash and most are designed to be held for long periods in order to see significant returns. There are a great deal of assets available to the investor, each of them requiring a substantial degree of experience in order to generate a solid return on investment:

  • Residential or Commercial Property
  • Government and Commercial Bonds
  • Equity Shares - Common Stock and Preferred Stock ("the stock market")
  • Investment Funds - Index Trackers, Exchange Traded Funds (ETF)
  • Collectibles: Art, Stamps, Cars, Wine
  • Private Small Businesses

When To Save and When To Invest

Unless you are part of the genetic lottery and have inherited a great deal of wealth, then in order to begin investing it is necessary to first begin saving. Once you have sufficient savings you will be in a position to begin purchasing investment assets.

Generate a cash liquidity buffer by keeping 9-12 months of monthly expenses in ready, liquid form. Leave this alone in a Cash ISA of high-interest savings account. Only then will you ready to begin investing.