Investment Outlook Essay for 2023

2022 Summary and Random Thoughts

Don't overestimate your ability to predict the future.

Looking back at the outlook for 2022 at the beginning of 2022 (2022 Investment Thoughts and Essays), it was found that although at that time it was considered to be as objective and impartial as possible, upholding value analysis and keeping a calm mindset, there were still many mistakes made, and the accuracy of the predictions was frighteningly low.

Our stock selection yield (calculated as full position, each product yield will be different) was around -20% (valued in local currency), which was similar to the Shanghai and Shenzhen 300 and S&P 500 -20% to -21%, better than the startup board and NASDAQ -30% and -34%, but worse than the Hang Seng Index -14%. However, excluding Hong Kong's sharp rebound in the past 1-2 months, the mid-year low actually reached 14,597, with a drop of nearly -40%. If it wasn't for the dramatic 11-12 months, Hong Kong might have been the worst market in 2022, with a situation as bad as Russia，which was sanctioned by the whole world...

In fact, in the past 100 years, situations like 2022, where almost all categories of assets have suffered such a sharp decline with the exception of a few, such as rising energy prices and the US dollar index, are rare. Many traditionally risk-hedged investment portfolios have failed, and it is extremely rare for even some traditional safe assets to suffer such huge declines, such as US long-term 20+ year government bonds, which fell by as much as -31%.

Figure: Performance of major global asset classes in 2022

Of course, there are many reasons we can blame, such as: higher-than-expected global inflation, significant interest rate hikes by the Federal Reserve, the Russia-Ukraine war, China's epidemic prevention policies, and so on. There are too many things that we have not seen in our shallow investment careers, and even very rare in history, making it difficult to form convincing generalizations. Even those investors who made correct judgments in the past year can hardly say they can repeat their success.

Image caption: The Federal Reserve's benchmark interest rate experiences massive fluctuations, far exceeding the expectations from the beginning of 2022.

Blind concentration in investments may be a form of overconfidence

The debate between concentrated and diversified investments is a widespread and enduring argument in the investment world, with each having very successful representatives, such as Warren Buffet being a typical example of successful concentrated investment. In fact, whether concentrated or diversified, investment is a wager against the market, for example, believing that one's chosen targets can achieve higher-than-average market returns. However, if all participants think this way, why would you be the ultimate winner? In fact, although the market may sometimes appear foolish in hindsight, its collective wisdom and pricing ability should never be underestimated. When deviating from the market, being highly concentrated, or even having high positions and leverage, one needs to be very cautious, as it is possible that the error lies with oneself.

Additionally, there is an important issue that even if the market is wrong, the time it takes to correct itself may far exceed your tolerance (or even the company's management's limit). In this case, blind concentration in investments may be a form of overconfidence.

How can we reduce the variance in the "batting average"?

Looking back at the past six years of our investment journey, the annual excess return rate before 2021 was 20%, and sometimes even close to 50%. However, in the recent two years from 2021 to 2022, we seem to have suddenly entered a stage that is not well-suited for our investment framework (value-based growth stock investing), with a significant decline in excess returns and negative returns in both 2021 and 2022, which is quite embarrassing.

The "batting average" (the probability of positive returns on holdings) is also a good indicator to measure our forecasting ability. In retrospect, the batting average of our positions in each phase, both long-term (assuming holding until now) and short-term (one-year holding period), has fluctuated significantly. This may be closely related to the investment environment of that year. For example, during the phase of rising from a low position, both long-term and short-term batting averages are quite decent. However, if the valuation is relatively high, the long-term batting average of holding until now becomes mediocre, and even at a very low position, the short-term batting average depends on the significant differences in market fluctuations that year. Sometimes the market environment is good, resulting in a higher batting average, and vice versa. Although we have always adhered to the value principle and carefully selected stocks, the actual results have shown significant differences across periods. Many times, it is difficult to show the difference in stock selection in the short term, and as the market fluctuates, the overall return of the portfolio is higher than the market at other times due to the outstanding performance of some star stocks (such as AMD, TSLA in certain stages causing differentiated performance in the portfolio), even if the batting average has not increased significantly.

In general, in the short term, macro and market conditions have a significant impact on the overall performance of the portfolio. We should not overestimate the ability of stock selection to outperform the market as a whole. The contribution of outliers in the portfolio may also lead to differentiated performance results compared to the market, even if the batting average during that period is not necessarily high.

Instead of predicting paths, focus on predicting probabilities

When we consider the reasons behind our positions at each stage (every month) in history, it certainly also includes our judgment of their future trends (whether based on game theory or various scenarios). However, the actual situation is that the probability of correctly predicting the path is unstable and depends on the actual situation at the time and the outcome of the investor group's game. The variables are too large, causing our batting average to fluctuate significantly.

At the end and beginning of each year, as well as at various special points in time, we can always see a variety of outlooks and predictions from both sellers and buyers. The results often have a low accuracy rate; people just selectively remember some of the magically correct or incorrect predictions. However, if we were to compile statistics, I believe the accuracy would not be high. In fact, every active fund manager makes various predictions about future paths when making allocations. Although some results are better than others, overall, active fund managers do not perform better than completely passive indexes (indexes are actually the combined result of all active management participants).

Although we know that no matter what position the stock price is in (overvalued or undervalued), there are only three possible future trend paths: rising, fluctuating, or falling. Therefore, predicting trends at this time is actually quite difficult because there are enough smart participants involved in pricing, and it is difficult to always be smarter than other participants in judging the subsequent trends. After all, the most difficult part of investing is not understanding the current situation but understanding the expected state and trend. For example, everyone knows that we are currently in a period of high inflation, but this fact itself does not have any effect on investment because it has already been priced. However, understanding where the expectations lie and how future inflation expectations will evolve, as well as how they will affect macro expectations, is a matter of differing opinions. Sometimes it's just cyclical fluctuations, and sometimes it's structural changes. Various explanations will emerge as events unfold (we are good at rationalizing because it makes us feel comfortable and consumes the least energy). In hindsight, everything always looks crystal clear and simple, but various explanations, even those that are later proven wrong, must have been very persuasive at the time (otherwise, so many smart investors wouldn't have believed them and reflected them in pricing).

Although it is difficult to understand the subsequent trends, especially to understand whether the various expectations of the entire market are sufficient, we still need to have a clear judgment on the long-term fundamental trends of the companies we research and invest in and their current valuations and market expectations.

In history, the periods when our medium and long-term batting average was high were often when the overall position was low (although it is difficult to understand whether the short-term trend will be lower or reverse) and when the selection of company fundamentals did not have significant judgment issues. At high levels (although we always like to use various reasons to justify the rationality of high valuations), we may continue to maintain high valuations, but once the valuation system collapses due to various reasons, our return rate and batting average will drop significantly.

Therefore, we say that even if it is difficult to understand the trend, we should at least try to understand the position and probability issues as much as possible.

Incorporate various non-base case scenario analyses

Looking back on our past investment experience, our pricing has been mainly based on a single base case, which we use to price the target company, calculate IRR, and finally determine the position by adding our confidence level.

Although this relatively simple analytical framework may try to be as objective and meticulous as possible when making judgments, we know that humans have a tendency for self-deception (our biases influence our thoughts), often unconsciously rationalizing decisions. Furthermore, the future is full of unknowns and uncertainties, and even with an objective and impartial view, judgments may still be incorrect. Investing is essentially making probabilistic judgments in the face of an uncertain future, which is inherently risky.

Relying on a single base case may inevitably lead to wishful thinking errors, imprecise judgments, and slow adjustments. We believe that a more systematic approach to market and company analysis is needed, incorporating both supporting and opposing perspectives and tracking changes that can confirm or refute the underlying assumptions. This approach involves conducting multi-case, multi-path research based on a core base case, and continuously adjusting the probability distribution of future paths based on actual developments.

Table: Some of our market logic assumptions and tracking models

For key individual stocks, we also have a similar system of pros and cons analysis and tracking.

Increasing Imagination for Nonlinear Event Occurrences

Due to the "energy-saving" effect in biological evolution, humans inevitably have a strong "inertia thinking," which is probably the most energy-saving thinking mode. Precisely because of this, human society is most likely to develop in the direction of inertia and continuity. Although inertia may be the case in most situations, due to the numerous influencing factors, there are always times when development trends enter turning points and fluctuations. Therefore, for investors (as well as official and authoritative judgments, which also often make mistakes), having only linear thinking logic may lead to many misunderstandings (of course, always guessing turning points might be worse).

Taking the U.S. monetary policy as an example, we can explain the judgment error problem for investors and government agencies. We cannot deny that both the buy-side, sell-side, and the Federal Reserve all have very powerful data, analytical tools, and analysis teams. However, a widespread phenomenon is that after the 2020 pandemic, the Federal Reserve flooded the market with liquidity, and the unanimous expectation was that the world might enter a long-term era of low or even negative interest rates. Even the great Dalio, the head of the world's largest hedge fund, said, "cash is trash"... However, as we know, the world (except for China) quickly entered a period of high inflation after the pandemic. Investors and authoritative institutions initially denied and resisted change, using various reasons to explain that inflation was "transitory." The expectation for rate hikes at the beginning of 2022 was only three modest 25bp hikes...

This inertia-denial-forced change-inertia thinking logic of humans seems to have not changed at all...

So, at this point in time, should we stick to inertia or fully consider its nonlinearity? Of course, this depends on the fundamentals and pricing.

At least historically and based on current market pricing, we can draw the following conclusions:

1. The Federal Reserve's policy seems to be full of dramatic actions and equally jaw-dropping shifts, which means that at many points in time, the Fed, with the most powerful resources in the world, lacks the ability to accurately predict the future (otherwise, there would be no need for such exaggerated actions like shooting up to the sky and then down to the ground).

2. From today's perspective, the U.S. economy and the Fed seem unable to maintain interest rates as high as 5%, well above the level supported by the long-term economic fundamentals, and sustain them for a long time (the Fed's own target is to fall back to 2-2.5% after resolving inflation issues).

3. The speed of the Fed's policy shift depends on the actual economic fundamentals (such as the speed of inflation decline, the speed of economic downturn, etc.), and the Fed's own interest rate dot plot is also very idealized and very controlled, with a slow decline. However, looking back at history, almost every actual effect is a "panicked" rapid adjustment of previous excessive interest rate policies. So what will happen this time?

4. Due to the Fed's repeated emphasis on inflationary pressures and investors' efforts to find lasting reasons, the current pricing is mainly for a slow decline in interest rates. If we calculate the pricing of long-duration U.S. assets (such as long-term government bonds), they are heavily influenced by short-term interest rates, and the underlying pricing of interest rate declines is even slower than the Fed's guidance. This expectation gap is actually significant.

Image: Historically, the Fed's actions have always been dramatic and "panicked." What about this time?

Outlook and Essays for 2023

As mentioned earlier, it is difficult for us to accurately predict the trends and developments for 2023 (the path). However, we can attempt to answer scenarios and probabilities of different outcomes (probability distributions of various paths) and analyze the current market pricing conditions.

The following are some of our possible guesses for various assets:

Domestic assets:

Especially Hong Kong stocks, although they have rebounded somewhat from the bottom, they are still in a stage of great divergence.

Bullish investors believe that the central government has completely shifted its focus to the economy, and the hurdle of epidemic prevention has already been overcome. With the gradual recovery of consumer and investor confidence, the economy is bound to stabilize and rebound. Currently, valuations and profit expectations are still at the bottom.

Bearish investors (interestingly, UBS downgraded its ratings on Chinese real estate on the first day of the year) believe that the Chinese economy is in a long-term downward cycle, with many long-term structural problems. What we are seeing now is just the beginning of a fall from a high point, far from a rebound, let alone a reversal.

Of course, there are also various other debates, including debates about geopolitical risks and whether the direction of marketization is appropriate. Bearish investors believe that the short-term rebound is only a temporary phenomenon in a long-term trend of structural bearishness.

We believe that all these arguments have their own logic and rationality, and the future direction of development is possible. However, in the short term (at least within a year), we believe that promoting the economy, promoting marketization, and enhancing confidence are the more probable directions. Despite various problems in the Chinese economy, the low base effects of excess savings and economic fundamentals and confidence over the past three years can produce a period of recovery. Of course, it is difficult to predict the duration and intensity of this recovery effect in advance, but in terms of current market pricing, we believe that there are still quite a few industries and companies (although not necessarily all, and not necessarily those that have already risen sharply and have expensive valuations in the market) that have considerable room for growth, giving us opportunities for stock selection.

Image: Has China's real estate market reached the peak of a long cycle and is it facing a long period of decline pressure, similar to the Japanese market from 1991 to 2009 and the US market from 2006 to 2012?

International assets:

1.Is the current pricing after a severe decline enough? When can we expect the fundamentals to bottom out?

Undeniably, current pricing has experienced a significant decline compared to 2021, and in the long term, this is a more attractive position. However, the core issue is whether our calculated IRR has overestimated its fundamentals and led to misjudgment.

As for the US stock market, the majority of sellers are bearish (more specifically, the sellers' strategy is bearish, but the bottom-up analysis of company analysts tends to be more optimistic), and the proportion of buyer positions is low (although compared to history, valuations may not be low enough). The bullish view is mostly based on long-term growth potential and the expectation that inflation and interest rates have peaked and may support a rise in valuations. However, the main problem is that the US economy may face relatively fast deceleration pressure under such high-interest rates, especially for some optional consumption (the most typical being credit consumption products such as housing and cars) or investment goods (including commercial and personal investment decisions) directly related to interest rates. With a substantial interest rate hike, the cost of such consumption will also increase significantly. In addition, the timing of purchasing such products is flexible and can be deferred or delayed, resulting in significant fluctuations in demand. In the stage of a substantial interest rate hike, the downward cycle's impact may be underestimated using previous linear analysis frameworks.

Especially for growth stocks, such impacts may be more severe. Take Tesla, which has experienced a significant decline in recent times, as an example. Even if its fans do not believe in Ma Yilong's 50% annual growth guidance, they must expect it to continue to provide compound growth of 20-30% and long-term options that could change the world (such as robotics technology and autonomous driving). When they suddenly discover that its book-to-bill ratio has fallen to 0.6, the new order's run rate is only 1 million vehicles, even lower than the delivery of 1.3 million vehicles in 2022, the shake-up of its long-term growth space's (whether the target is 5 million or 20 million vehicles) must be a psychological collapse. In this case, its pricing will shift towards the logic of the short position. For companies with particularly polarized pricing, this can easily lead to a significant decline.

Of course, for long-term investors, their confidence in the base case's target may still depend on many underlying company operational capabilities. They may think that the current demand issue is only short-term, and it is still possible to return to growth in the future with continued cost-down and brand influence improvement. However, for such arguments, we also need to continuously track evidence to find support, rather than blindly believing it based on wishful thinking.

Overall, we still believe that the long-term potential of US growth stocks is enormous. However, during the short-term economic downturn stage, there may be challenges to the expectation of long-term space and profit-making ability, such as the previously huge space and profit-making ability only being caused by short-term non-sustainable factors. For some companies with proven long-term growth potential, we believe that there is a high probability of broad upside potential after facing challenges in 2023.

Of course, for US bonds, especially long-term Treasury bonds, we do not need to consider default risks but only need to focus on whether long-term inflation expectations can return to the normal 2% level to support the current pricing of long-term interest rates as high as 4%. Based on our previous analysis of the US economy, inflation, and interest rates, we believe that the current odds and win rates are sufficient.

2.When expectations are low enough, is the possibility of exceeding expectations increasing?

Time is an interesting thing, and an important aspect is that expectations gradually change over time, and these changing expectations in turn change the fundamentals step by step. In addition to the superficial impact of declining valuations due to investor expectations, adjustments in time and expectations also gradually change the decisions of companies and all market participants. For example, when pessimistic expectations prevail, on the one hand, consumer decision-making becomes more cautious, putting pressure on business revenues. On the other hand, dealers and companies reduce inventory, capital expenditures, recruitment, and even layoffs or bankruptcies, etc. Such conservative decisions during a downturn may lead to further contraction pressure as conservative decisions among different stakeholders can have a cumulative effect, further exacerbating the cyclicality. However, when the cycle gradually bottoms out (which is essentially when the expectations of all decision participants reach a bottom), and gradually enters an upward cycle, many additional upward elasticities are found. In addition to consumer confidence recovery, businesses' defensive measures during the downturn, such as lower inventory, higher personnel efficiency, and fewer competitors, will bring brighter prospects for their future.

Therefore, for 2023, an important question is whether, when everyone is expecting a recession in the US economy (as we do), the long-term optimistic expectation of an acceleration in growth and a recovery in profit margins will emerge in the second half of 2023 or 2024 after the end of rate hikes, or even after rate cuts?

Of course, perhaps we are proposing this hypothesis a bit early because the US economy seems to have just begun to suffer from the impact of the severe rate hikes in the second half of 2022. However, expectations are an interesting thing, and when everyone clearly sees the "light at the end of the tunnel," the transformation may be very fast. This is just like how we are still in the peak period of the domestic epidemic, but everyone can clearly envision the upward economic cycle after the epidemic recovery, and thus will not assign long-term pricing to the current short-term crisis.

One of the most difficult things about investing is that it can often be hard to have a clear point that tells you how far the market has gone. Having just enough expectations is not enough, and having too much is not enough, especially since most participants are quietly conducting their own calculations, and the only result that can be publicly displayed is the price itself.

What is openly told to you is only the seller's prediction, and these predictions are sometimes similar to the buyer's (many buyers also refer to the seller's views), but there are also sometimes significant differences. The seller's prediction itself may only be a slow variable, and in the minds of many buyers and market participants, many expectations have actually been adjusted in advance. Over the course of 50 years, seller predictions and the trend of the year often do not match.

Oaktree's summary of the differences between the current and the past 12-year cycle shows that the pricing cycle that began in 2022 is completely different, and it is difficult for us to accurately know how long it will continue, but it clearly brings us very different challenges and opportunities.

Finally, we also list the pricing situation of various assets that we calculate. This reflects our profit forecast under our base case scenario, but due to our limited ability, there may also be deviations.

Note: The calculation model is based on many assumptions and is not intended as investment advice. It is only for reference. In addition, in some of our previous articles, we have also expounded on some of our macro analysis and framework for reference: Some thoughts on investment methodology and pricing issues, some thoughts on investment area selection, some thoughts on exchange rate issues, some thoughts on interest rate issues, some thoughts on inflation issues.