Friday, January 20, 2023

On Timing, Sizing and Sell Discipline

Today, let’s spend some time to talk about some of the important nitty gritty of investing. This is not the usual fun and games - big ideas, cheap value names or best-in-class companies. Everyone loves ideas, what to buy, deep dives. But what actually brings the dough home is good money and portfolio management. As such, we need to talk about sizing and timing and how we should construct our investment portfolio.

There are three crucial aspects: sizing, timing, capacity and a fourth all-important factor: sell discipline. A lot has been said about how an investment will make money when we buy at the right valuation, but real money only comes into our pockets after we sell. So let’s talk practical about that too. First, it is about sizing the bet.

1. Sizing

Most people do not think too much about sizing and I seldom read literature about sizing which is unclear why given its importance. Perhaps sizing reveals too much financially or maybe we assume everyone knows how to size? But here is what I have figured out over the course of my investment career. Sizing wrongly hurts a lot. It is not easy, it requires practice, everyone has different thresholds and hence we should discuss seriously about sizing.

The first question to ask is how much can you lose and not be affected psychologically? Is it $10k or $100k? Or maybe it is lower or higher. There is no shame about it. If it is $1,000. Then that is your maximum bet size. Don’t mess with your mind. If losing $1,000 makes you unable to sleep, what is the point of playing this game with higher stakes and hurting yourself emotionally and psychologically?

So start with the size that makes you comfortable.

There is a related question which would be if we only bet $1,000, how can we get rich? Well, it will take more than a few ten-baggers and home-runs, but thankfully, low commissions today can make $1,000 bets go far. Back in the days when there is a minimum commission of $20 per trade, it was not feasible to bet $1,000 because you incur 4-6% of transaction cost just by buying and selling. But today, we can do it!

For more practical numbers, let’s use something with some macro-economic basis. I would start with $50,000. That is my maximum bet (not to be confused with initial bet) and a certain delta around that number might work better for you. Why $50,000? Well, it was slightly higher than my first annual paycheck and it is also in the same ballpark as the GDP per capita of OECD countries, which is $42,000 according to the link below.

https://data.worldbank.org/indicator/NY.GDP.PCAP.CD?locations=OE

But let’s talk about first paychecks. That’s more interesting. When I started work about two decades ago and took my first paycheck home. I was so happy, I gave money to my parents. I have enough money for the first time to afford stuff. It feels good. Everyone remembers their first paycheck.

So when I lost on an investment that was bigger than this first annual paycheck. Imagine the pain. Imagine I had to tell my better half I lost that much money. I couldn’t sleep just thinking about it. So that was how I figured out, my maximum loss is $50,000. It doesn’t matter if you think you have done your homework. You know the stock or investment and you are sure it will not drop beyond 50% or 80% therefore using this way to think about a maximum bet size is wrong. No it is not because if you have invested enough, one of those wrong bets will go to zero.

So when you have determined that number. Make sure you don’t ever buy more than that in one investment. Next, we need to build that up over tranches. You don’t put all $50,000 on Day 1 because you will never know if the stock will go down more and you lose the chance to buy at a lower price.

My rule of thumb is to think in baseball batting terms. You have three swings. After that, you are strikeout. Each swing you place 1/3 of the position. Some people like to do more, some less. It depends on how well you can do this, it is a subtle art.

Some people say it means a lack of conviction. If you are sure, just go all in. They have never invested. You go all in, you can get strikeout after the first swing. How does that feel? If you are really confident, maybe you can go 1/2 instead of 1/3 with your first swing. Going all in doesn’t end well most of the time. Trust me, been there, done that. Nope. Didn’t go well.

So if the first swing turns out well, the stock runs, then unfortunately, you cannot deploy the full amount. At least we have benefited. But if it didn’t, this is when the second and third swings will count. Here we will need time diversification.

2. Timing

Timing is about time diversification. You would have done a lot of homework before the first swing. So by and large, there will be enough upside. You know the margin of safety. But to make this work better, say the first swing didn’t go as planned and the stock corrected 10%. Then it pays to wait one month and take advantage of market movements later for the second swing. Of course, you also want to pay attention to the price. If it drops between 10-15% lower, then it is good to get in by averaging down.

There is another reason to think monthly or even longer. We are all busy with our lives, if this is not your day job, maybe spending a few hours once a month to focus and think and then execute is the best option. Don’t go buying today, buy more tomorrow if it drops or sell next week after you made 5%. It takes up too much energy. So set aside a time every month to think and trade. Then move on, come back and monitor monthly.

There are exceptional times when you need to do a lot in a few days. Think March 2020. Pandemonium struck but it was also the best time to buy. It takes guts. You have be able to recognize such times and deploy money well. Most investors will not be able to do so. They have either lost too much to think straight or just scared cold and unable to move. I would say you need to think in terms of your portfolio, not individual bets. Ideally, if you can deploy 50% of your portfolio in Mar 2020, you would have created a huge positive impact. There is no second chance next month. The window will be just days. But it is so scary that it is hard to move a lot. You are figuratively catching a falling knife with your bare hands, maybe even your kids’ hands. So try your best. Test your capacity.

3. Capacity

Besides the capacity of your gut i.e. ability to take losses, capacity is also about the number of bets, stocks, investments, ideas that you want to have. Most laypeople think that a portfolio, especially an individual or retail portfolio should have just a handful of bets. Depending on the individual it could be 10 or 15 bets at the maximum. For some people, it could be just 5 or 6 concentrated bets.

But it comes back to sizing. Unless you can stomach big losses, having say 5 bets, each bigger than GDP per capita of OCED countries is not something everyone can do. Hence a lot more bets at your maximum bet size makes sense. The upper bound could be your capacity to monitor. If you don’t want to monitor more than 10 bets. Then it is 10.

However to enjoy the benefits of diversification, one of the few free lunches in investing, maybe the number of bets should be big. The CFA textbook says it should be 30. But most people may think that is too much diversification. They cannot remember yesterday’s lunch, or 3 things the last writer asked them to remember, let alone 30.

In statistics, recall that we learnt about the Law of Large Numbers. So what is the smallest number that we can to be considered a Large Number? Remember N? Our teacher Mrs Shirley would say N >= 30. So maybe it should still be closer to 30.

Some astute investors don’t subscribe to this. Except for Peter Lynch, who managed Fidelity’s Magellan Fund, became one of the most celebrated successful portfolio manager and he held over 1,000 names. Today, most good investors think that an ideal portfolio should have 10 to 15 names. Warren Buffett said it was 20 for him. So, the idea is to pick your 10-20 best ideas and rake it in. You cannot have 30 best ideas, surely some of them are not best by the time you get #29 on that list.

I do not know which school is right? I do like to stick to the textbooks, maybe it is good to try to get 30. What worked for me is to have 8-10 top ideas and another 8-10 potential top ideas and the last few tail ideas that you want to have it in case they become so big for whatever reasons. This will be the goal for this newsletter. We will aim to get to 30 ideas!

4. Sell Discipline

When we have our best ideas, we were taught to buy and hold forever. “Our holding period is forever.” so says the Oracle of Omaha. No value investors talked about sell discipline. Naturally, I was brought up to think buy and hold.

It didn’t work for me.

Selling is so important. To sell well is perhaps the hardest part of the art. Well, to be fair, Warren Buffett did say never sell the best names and if you must, only sell when:

1. you have a better opportunity

2. you need the money for more urgent matters

3. when the investment thesis has gone wrong or has changed

I am not about to refute the Oracle so the above reasons are definitely the best reasons to sell. But as alluded throughout the article, investment is about timing, sizing and the many nitty gritty important details. So, the big idea is that we don’t have to sell everything all at once. We only sell 1/3 or 2/3 and we can hold the remaining 2/3 or 1/3 forever. We can use time diversification to sell over time when the right reasons present themselves to sell. With this in mind, I have added three more reasons when we should sell.

1. We should sell when valuation is rich → sell 1/3 or 1/2 depending on how expensive the stock has become. We can then recycle that capital to better names, which is #2 below.

2. We should sell to rebalance the portfolio (for me this means keeping bets at $50k if they have grown to $100k, ie lucky me! But we need to bring the notional sum down because losing all that gains back is very bad psychologically.

3. Lastly, we should sell when markets are overall expensive (e.g. Dec 2021). Although it is difficult to do so because at that point in time, we won’t know whether it is at the peak. As such, we diversify the selling, ie selling 1/3 or 1/4 or in different proportions. It is also good to accumulate dry powder during such periods so that we can deploy back when markets turn cheap.

It is important to understand your own style as well. If you tend to be too early when buying, then buy slower with the first bets being smaller accordingly. If you tend to overstay, then start doing bigger first sells. Understand your maximum loss and size according to your most comfortable level, diversify both the number of names and buys / sells over time, figure out your best ideas and structure the portfolio accordingly. When it is time to sell, look closely at valuations and the overall markets and document everything. With these steps hopefully we can get the portfolio to grow well. Target just 2-3% of value up every month, over time it will compound crazily.

Huat Ah!

Thursday, January 05, 2023

Books #18: Security Analysis - Part 3

In this last post on Security Analysis, we would discuss some of the difficulties analyst faced when analysing stock then and now. The authors have put some thoughts out elegantly which is worth studying here. Here are a few lines from the book, including one of the most famous line in full and in bold:

The market and the future present the same kind of difficulties. Neither can be predicted or controlled by the analyst, yet his success is largely dependent upon them both. The major activities of the investment analyst may be thought to have little or no concern with market prices.

The market is not a weighing machine, on which the value of each issue is recorded by an exact and impersonal mechanism, in accordance with its specific qualities. Rather should we say that the market is a voting machine, whereon countless individuals register choices which are the product partly of reason and partly of emotion.


Alongside the famous line on the marketing being a voting machine, the authors also lay out this chart above to try to decipher what goes into determining market price. In short, there are so many things that make prices unpredictable in the short-term but yet it is the stock analyst's job to understand everything because in the long term, analysis of the business and its earnings power makes the difference.

Short term prices are also inexplicable but Wall Street wants actions and will explain for it, regardless whether the explanations are justified. This remains one of the hardest job of the day even for seasoned investors - to explain price action fo the day. Sometimes, stocks jump 5% for no reason and we only figure it out over the next few days.

Again, here are lines from the authors themselves:

The exaggerated response made by the stock market to developments that seem relatively unimportant in themselves is readily explained in terms of the psychology of the speculator. He wants action, first of all; and he is willing to contribute to this action if he can be given any pretext for bullish excitement (whether through hypocrisy or self-deception, brokerage-house customers generally refuse to admit they are merely gambling with ticker quotations and insist upon some ostensible reason for their purchases.) 

Stock dividends and other favorable developments of this character supply the desired pretexts, and they have been exploited by the professional market operators, sometimes with the connivance of the corporate officials. The whole thing would be childish if it were not so vicious. The securities analyst should understand how these absurdities of Wall Street come into being, but he would do well to avoid any form of contact with them.

As such, in order to profit from stock investment, we come back to valuation. We cannot win trying to profit from short term market movement. The only logical way to win is to determine the intrinsic value of the stock and buy with sufficient margin of safety. The following paragraph, the authors cautioned that general market conditions ie beta sometimes overwhelm everything so we need to be careful.

Investment in bargain issues needs to be carried out with some regard to general market conditions at the time. Strangely enough, this is a type of operation that fares best, relatively speaking, when price levels are neither extremely high nor extremely low. The purchase of "cheap stocks" when the market as a whole seems much higher than it should be will not work out well, because the ensuing decline is likely to bear almost as severely on these neglected or unappreciated issues as on the general list. 

On the other land, when all stocks are very cheap, there would seem to be fully as much reason to buy undervalued leading issues as to pick out less popular stocks, even though these may be selling at even lower prices by comparison.

On valuations, it was refreshing to see that the authors did lay out something concrete and also attributed the development to a certain Roger Babson (pic above), whom I didn't know who he was but apparently someone famous back then. The following is his rule of thumb:

The multipler might be equivalent to capitalizing the earnings at twice the current interest rate on the highest grade industrial bonds. The period for averaging earnings would ordinarily be seven to ten years.

In short, if the bond yields are at 3-4%, then earnings yield should be double of that at 6-8% and we should use the averaging earnings over 7-10 years. So we come back to why we should always buy stocks trading at teens price earnings, which can be difficult in today's context, so I would carefully stretch that to low twenties. 

Last but not least, here's their words on wisdom on trading. In trading, there is no margin of safety, you are either right or wrong and if you are wrong you lose money.

The cardinal rule of the trader that losses should be cut short and profit safeguarded (by selling when a decline commences) leads in the direction of active trading. This means in turn that the cost of buying and selling becomes a heavily adverse factor in aggregate results.